Professional Documents
Culture Documents
High Penetration of PV in Local Distribution Grids REPORT PVPS T14 02 2014 PDF
High Penetration of PV in Local Distribution Grids REPORT PVPS T14 02 2014 PDF
P
The International Energy Agency (IEA), founded in November 1974, is an autonomous
body within the framework of the Organisation for Economic Co operation and
Development (OECD) that carries out a comprehensive program of energy cooperation
among its 23 member countries. The European Commission also participates in the work of
the agency.
The )EA Photovoltaic Power Systems Programme )EA PVPS is one of the collaborative
research-and-development agreements established within the IEA, and since 1993 its
participants have been conducting a variety of joint projects in the applications of
photovoltaic (PV) conversion of solar energy into electricity.
The mission of the IEA-PVPS program is to enhance the international collaborative efforts
that facilitate the role of PV solar energy as a cornerstone in the transition to sustainable
energy systems by
1. Ensuring sustainable PV deployment,
2. Improving PV performance and reliability, and
3. Assisting in the design of new market structures and regulations that will be suitable for
the widespread adoption of unsubsidized PV.
The overall program is headed by an executive committee composed of one representative
from each participating country, whereas the management of individual research projects
(tasks) is the responsibility of the operating agents.
The overall goal of the IEA-PVPS Task 14, (igh Penetration of PV Systems in Electricity
Grids, is to promote the use of grid connected PV as an important source in electric power
systems at the higher penetration levels that may require additional efforts to integrate
dispersed generators. The aim of these efforts is to reduce the technical barriers to
achieving high penetration levels of distributed renewable systems.
The current members of the IEA-PVPS Task 14 include Australia, Belgium, Canada,
Switzerland, China, Germany, Denmark, Spain, Israel, Italy, Japan, Portugal, Sweden, and the
United States of America.
The IEA-PVPS Task 14 is organized in subtasks. Figure 1 shows the organizational
structure of this international collaboration.
iii
2Austrian
BELGIUM
Karel De Brabandere3 and Carlos Dierckxsens3
33E,
Brussels, Belgium
CHINA
Wang Yibo4
4Chinese
GERMANY
Thomas Stetz,5 Markus Kraiczy,5 Konrad Diwold5
5Fraunhofer
GREECE
Stathis Tselepis6
6Center
for Renewable Energy Resources and Saving (CRES), Photovoltaic Systems and
Distributed Generation Department, Athens, Greece
ITALY
Adriano Iaria,7 Antonio Gatti,7 and Diego Cirio7
7Ricerca
sul Sistema Energetico RSE SpA, Energy Systems Development, Milano, Italy
JAPAN
Yuzuru Ueda,8 Kazuhiko Ogimoto,9 and Koji Washihara10
8Tokyo
10New
11European
SWITZERLAND
Davy Marcel12 and Christof Bucher13
12PLANAIR
13Basler
vi
A
The Australian authors would like to acknowledge the contributions of their CSIRO and
University of New South Wales UNSW Colleagues towards Australian participation in this
PVPS Task, funding support from the Australian Renewable Energy Agency and industry
contributions from members of the Australian PV Institute.
The Austrian contribution to this work was supported by the Federal Ministry of
transportation, innovation and technology (BMVIT) and the FFG within the framework of
the Project "HiPe-PV" (FFG Nr. 827762)
The Belgian contribution to this work has been funded by the three Belgian regions:
Brussels Hoofdstedelijk Gewest/Rgion Bruxelles Capitale, Vlaams Gewest and Rgion
Wallonne
The German contribution is supported by the German Federal Ministry for the
Environment, Nature Conservation, Building and Nuclear Safety and the
Forschungszentrum Jlich Gmb( PTJ within the framework of the Project (iPe-PV
(FKZ 0325266).
The Italian contribution has been financed by the Research Fund for the Italian Electrical
System under the Contract Agreement between RSE S.p.A. and the Ministry of Economic
Development - General Directorate for Nuclear Energy, Renewable Energy and Energy
Efficiency - in compliance with the Decree of March 8, 2006.
The U.S. contribution to this report was supported by the U.S. Department of Energy under
Contract No. DE-AC36-08-GO28308 with the National Renewable Energy Laboratory.
vii
viii
ix
is radial. ...........................................................................................................................................................81
Figure 32: Total net present value (NPV) of investigated local voltage control strategies
referred to the beginning of year t1. NPVInvest: Grid reinforcement, NPVOP1: Network losses
+ maintenance, NPVOP2: reduced PV feed-in [84] ..............................................................................85
Figure 33: Measured and simulated power flow at a HV/MV substation (left) and reactive
power fluctuations at a HV/MV substation for different local control strategies of PV
inverters (right) on a summer day with alternating solar irradiation [90]............................86
Figure 34: Influence on Q-Exchange by changing power factor of LV PV units and
optimization potential by applying the central control strategy on a summer day with a
clear sky [92]..................................................................................................................................................88
Figure 34: Results of the annual active and reactive power flow (1/4-hour mean values) at
the TSO connection point with central control (green crosses) and without central
control blue crosses for a cos of
of the LV PV units
................................................88
Figure 36: Annuity of investment plus annual operational costs for the scenario with a power
factor of 0.95 (lagging) of the LV PV systems and for different PQ-bands at the connection
point to the TSO [93] ...................................................................................................................................89
Figure 37: Annual and cumulative installed PV capacity in Greece (estimated for 2013) .......92
Figure
Generation mix for Crete s power system on May
and RES penetration
level (violet line) [99] .............................................................................................................................. 101
Figure 39: Percentage of installed PV plants in the Italian regions (2012) [106] .................... 105
Figure 40: Hosting capacity in each MV node. Cumulative histogram of MV node percentages
with the concerning installable DG powers [117] ......................................................................... 110
Figure 41: Hosting capacity in each LV node. Cumulative histogram of LV node percentages
with the concerning installable DG powers [118]. Maximum fast-voltage deviations 5%111
Figure 42: Hosting capacity in each LV node. Cumulative histogram of LV node percentages
with the concerning installable DG powers [118]. Maximum fast voltage deviations 10%111
Figure 43: General structure of the Japanese power system ............................................................ 115
Figure 44: Frequency feedback function of JSIDF ................................................................................. 121
Figure 45: FRT requirement for PV systems........................................................................................... 123
Figure 46: Schematic of the supply demand balance throughout 24 hours in a system with a
large number of PV systems .................................................................................................................. 124
Figure 47: A typical distribution network with various voltage control equipment ............... 125
126
Figure 48: Configuration of ADAPS with LPC and SDI ......................................................................... 126
Figure 49: Loop controller setup and results ......................................................................................... 127
Figure 50: Evaluation results of proper voltage regulation methods according to DG
penetration rate ......................................................................................................................................... 128
Figure
Flow chart of developed customer s appliances operation planning method ...... 129
Figure 52: An example of demonstration results of a developed operation planning method130
Figure 53: Overview of the networks considered in the case studied: a MV grid (left part of
the figure) and a HV grid (right part) (Source: ReserviceS project 2014 [135]) ................ 133
Figure 54: High-level description of the methodology and assumptions used for the analysis
(Source: ReserviceS project 2014 [135]) .......................................................................................... 136
Figure 55: Scope of the analysis (Source: ReserviceS project 2014 [135]) ................................. 137
Figure 56: Results of the voltage violations analysis for the R/X2 MV network (Source:
ReserviceS project 2014 [135]) ............................................................................................................ 138
xi
Figure 57: Impact of voltage control on active power losses (Source: ReserviceS project 2014
[135]) ............................................................................................................................................................. 139
Figure 58: Effectiveness of Q-V control depending on the R/X ration and the scenario
(Source: ReserviceS project 2014 [135]) .......................................................................................... 139
Figure 59: Number of occurrences and volume (in MVar) of unfulfillments (Source:
ReserviceS project 2014 [135]) ............................................................................................................ 140
Figure 60: Active power loss variation for the different considered scenarios (Source:
ReserviceS project 2014 [135]) ............................................................................................................ 141
Figure 61: Reactive power generation for the different considered scenarios (Source:
ReserviceS project 2014 [135]) ............................................................................................................ 142
Figure 62: Overview of Swiss import/export and transmission in 2012 (Source: Swissgrid)146
Figure 63: Overview of Swiss transmission and distribution grid (Source: Swissgrid) ......... 147
Figure 64: Structure of the documents for the liberalized market rules (Source: VSE/AES,
Swiss distribution assessment) ............................................................................................................ 149
Figure 65: P requirements applicable to transmission grid (Source: chapter 6.5.2 from TC
2010). ............................................................................................................................................................. 152
Figure 66: P delivered to be guaranteed from p power source connected to transmission grid
(Source: chapter 6.5.2 from TC 2010). ............................................................................................... 153
Figure 67: Main figure for frequency control (Source: chapter 6.5.5 from TC 2010). ............. 154
Figure 68: Grid topology of the investigated grid area Luchswiesenstrasse ........................... 155
Figure 69: Voltage histogram and maximum current loading for different PV penetration
scenarios....................................................................................................................................................... 157
Figure 70: PV energy production in Switzerland since 1990 (Source: Swissolar) .................... 158
Figure 71: Typical simplified electrical power system structure in the United States ........... 161
Figure 72: Typical simplified flow diagram of the interconnection process of interconnecting
DG .................................................................................................................................................................... 164
Figure 73: Circuit diagram of the NREL/SCE High-Penetration PV )ntegration Project s
Fontana, California, study circuit......................................................................................................... 166
Figure 74: Surface graph of POI voltage change for a range of operating power factor set
points and for a range of power production loss metrics ........................................................... 168
Figure 75: Single-line diagram of the IEEE 34 node test feeder showing the location of the
two voltage regulators and two fixed-shunt capacitors .............................................................. 171
Figure 76: Color graphs of the POI voltage (node 840) for every minute of the year in 2010
with a) no PV interconnected and b) a 2-MW PV system connected at node 840 .............. 171
Figure 77: Model results of voltage regulator operation during periods of highly variable
generation on June 16, 2010. The graphs on the right show regulator operation during
the entire day. The graphs on the left show operation from 10 a.m. to 1 p.m. .................... 172
Figure 78: The modeled voltage regulator operation impact of 1 MW of interconnecting PV at
any point along the distribution circuit ............................................................................................ 173
Figure 79: Applied Q(V)/P(V) characteristic in the German case study [84] ............................. 190
xii
Table 1: Summary of Australian Statistics for Solar PV Electricity Generation and Demand.... 6
Table 2: Australian Categorization of Nominal Voltages ........................................................................ 7
Table 3: Small-Scale PV Generation (Less Than 100 kW) in Australia in November 2013 ........ 8
Table 4: Statistics on Australian DNSPs......................................................................................................... 9
Table 5: System Voltage Requirements Under AS6100.3.100 .............................................................16
Table 6: Voltage and Frequency Requirements for Australian States and Territories ..............17
Table 7: System Size Limits and Voltage and Frequency Inverter Set Points ................................18
Table 8: Summary of Key PV Penetration Experiences/Issues on the Alice Springs Network 23
Table 9: Summary of the Key Findings of the Carnarvon Case Study ...............................................25
Table 10: PV Penetration Levels on Magnetic Island..............................................................................30
Table 11: Summary of the Experiences of High PV Penetration in Magnetic Island and
Townsville .......................................................................................................................................................31
Table 12: Demand Response Modes for Inverter Performance .........................................................37
Table 13: Statistics on PV Generation and Power Consumption in Austria ...................................40
Table 14: Requirements for Reactive Power Control for PV Systems ..............................................42
Table 15: Compensation of the Voltage Rise for Cable and Overhead Lines (for cos=0.90) ..49
Table 16: Voltage Drop Over the Transformer Main Reactance (for cos=0.90).........................50
Table 17: Statistics on PV Generation and Power Consumption in Belgium .................................55
Table 18: Statistics on PV Generation and Power Consumption in China ......................................65
Table 19: Basic Information on HV Level ....................................................................................................66
Table 20: Basic Information on MV Level ...................................................................................................66
Table 21: Basic Information of LV Level......................................................................................................67
Table
Requirements for Voltage Deviation in China s Distribution Grids According to
GB/T 12326 ....................................................................................................................................................68
Table 23: Statistics on PV Generation and Power Consumption in Germany................................74
Table 24: Basic Information on HV Level ....................................................................................................76
Table 25: Basic Information on the MV Level............................................................................................78
Table 26: Basic Information on the LV Level .............................................................................................80
Table 27: Requirements for Voltage Deviation in German Distribution Grids According to EN
50160 ................................................................................................................................................................82
Table 28: Requirements for Active Power Curtailment for PV Systems ..........................................83
Table 29: Requirements for Reactive Power Control for PV Systems ..............................................83
Table 30: Investigated Control Strategies of the HV/MV Voltage Transformer and PV
Inverters in [90] ............................................................................................................................................86
Table 31: Statistics on PV Generation and Power Consumption in Greece at the End of
December 2012 .............................................................................................................................................93
Table 32: Voltage and Frequency Operation Limits ...............................................................................97
Table 33: Compatibility Levels for Maximum Harmonic Distortion in LV and MV Networks .98
Table 34: Statistics on PV Generation and Power Consumption in Italy ...................................... 104
Table 35: Requirements for Active Power Control of PV ................................................................... 108
Table 36: Requirements for Reactive Power Control of PV .............................................................. 109
Table 37: Statistics on PV Generation and Power Consumption in Japan.................................... 114
Table 38: Outline of the Technical Requirements Guideline of Grid Interconnection to Secure
Electricity Quality in Japan (the criteria applicable only to PV) 1; *1: Specific case denotes
xiii
that the capacity of the inverter is less than 2 kVA in 1p2w, less than 6 kVA in 1p3w, and
less than 15 kVA in 3p3w, or that the power factor at the point of common coupling is
usually close to 1, such as in residential systems. *2: ex. single-phase two-wire)............. 117
Table 39: Outline of the Technical Requirements Guideline of Grid Interconnection to Secure
Electricity Quality in Japan. (The criteria is applicable only to PV.) 2 ................................... 118
Table 40: Outline of Grid Interconnection Code (JEAC9701). *1. Standard active islanding
detection scheme: A frequency feedback method with step reactive power injection (JEM
1498 (2012 JEMA)). See 2.8.3. *2. Voltage Q (reactive power) control .................................. 119
Table 41: Statistics on PV Generation and Power Consumption in Spain .................................... 132
Table 42: Characteristics of the Two Networks Considered for the Analysis. (Source:
ReserviceS project 2014 [135]) ............................................................................................................ 134
Table 43: Wind and PV Scenarios Envisaged for the Analysis of the Voltage Control in the MV
Network (Source: ReserviceS project 2014 [135])........................................................................ 134
Table 44: Wind and PV Scenarios Envisaged for the Analysis of the Voltage Control in the HV
Network. (Source: ReserviceS project 2014 [135])....................................................................... 134
Table 45: Results of the CBA Methodology Applied to Two Different Network Conditions.
(Source: ReserviceS project 2014 [135]) .......................................................................................... 143
Table 46: Facts and Figures of the Grid .................................................................................................... 155
Table 47: Results Summary of the Case Study ....................................................................................... 156
Table 48: PV Power Output Variation Expected for Modeled Year 2011 ..................................... 169
Table 49: Magnitude of PV Systems Point of Interconnection Voltage Change Related to
Operating Power Factor Set Point and Expected Production Power Loss (Red Numbers
Indicate Voltage Variation Higher than 0.7 volts on a 120-V base.) ....................................... 169
Table 50: Applied Local Control Strategies in German Case Study [84] ....................................... 190
xiv
1 I
This report presents the International Energy Agency Photovoltaic Power Systems
Programme )EA PVPS Task 14 Subtask 2 case study collection on high penetrations of
photovoltaics (PV) in local distribution grids. National experts from industries and
research in 11 countries present the current status of PV penetration as well as the latest
findings from their home countries regarding the technically and economically improved
distribution grid integration of PV. A management summary of the case study collection can
be found in:
Transition from Uni-Directional to Bi-Directional Distribution Grids
Recommendations Based on Global Experience. Management Summary of IEA Task
14 Subtask 2.
1.1 Motivation
When the IEA Task 14 was founded in 2010, annual PV installation rates were at their
maximums in many countries throughout Europe. The vast and rapid PV deployment led
and still leads to many technical and economic challenges on distribution systems
because the behavior of some distribution grids changes from one of consumption to one of
supply. Local distribution system operators were forced to instantaneously react to the
locally growing PV penetration rates with revised grid planning and grid operation
processes that, until then, had been tailored to serve consumption grids only and were well
established during the past decades. In some countries, these processes were also fueled by
a strict regulatory framework that limited the technical possibilities for distribution system
operators to cope with the rising installation rates in a technically effective and
economically efficient manner (e.g., the primary grid access for renewable energies as
defined by the German Renewable Energy Sources Act). As a result, many different
research projects were initiated on national levels that aimed to review grid planning and
grid operation processes to improve the technical as well as economic integration of PV in
distribution grids.
The motivation of IEA Task 14 Subtask 2 is to bring national experts on distribution grid
integration of PV together to share and discuss the latest research results and national
developments dealing with high PV penetrations in local distribution grids. This
international collaboration
Opens access to the latest market developments and changes within the respective
national regulatory frameworks of the participating countries (Information); and
A management summary of the overall IEA Task 14 Subtask 2 activities can be found in:
Transition from Uni-Directional to Bi-Directional Distribution Grids
Recommendations Based on Global Experience. Management Summary of IEA Task
14 Subtask 2.
The authors of the respective national reports are solely responsible for their content.
2 H
D
P
C
Australia
Authors:
Ben Noone,1 Anna Bruce,1 and Iain MacGill1
1Centre
Using these definitions, there are slightly more than 765,000 km of distribution lines and
cables in the major Australian electricity systems. Of this total, 14% are underground, 28%
are classified as low voltage (640 V and below), and 22% are single-wire earth return
(SWER) [10]1. Data does not exist for the distribution of PV systems across different
voltage levels; however, almost all solar PV capacity in Australia is composed of systems of
less than 10 kW connected to the low-voltage (LV) distribution network (see Figure 2).
This means that the penetration of PV in relation to the proportion of customers who have
PV systems is high compared to most other countries. For example, the percentage of
customers in the state of South Australia who have a PV system is approaching 20%
(Table 3).
SWER lines are low-cost, single-phase distribution lines that utilize a single wire and an earth-return path (no
neutral). SWER lines are commonly used in remote, sparsely populated parts of Australia, notably Queensland, and
typically operate at voltages of 12.7 kV or 19.1 kV.
Table 3: Small-Scale PV Generation (Less Than 100 kW) in Australia (status: 11/2013).
State or Territory
Number of
PV
Installations
13,426
248,085
2,664
353,622
154,194
17,177
196,478
147,185
1,132,860
Installed
Capacity (MW)
Customers with
PV (%)
37
613
10
983
438
51
509
335
2,976
7.9%
6.5%
3.1%
17.6%
18.7%
6.2%
7.5%
13.1%
-
Source: Clean Energy Regulator (status 4/11/2013); Australian Energy Regulator 2012
Company
Ownership
Number of
Customers
Km Line
RAB
(2010
$m)*
635
Maximum
Demand
2010 11
(MW)
701
Customer
Density
(Per km
Line)
34.3
ACT
ActewAGL
168,937
4,922
NSW
AusGrid
50/50
Government/private
Government
1,619,988
49,781
8,965
5,812
32.5
NSW
Endeavour
Government
877,340
34,172
3,925
4,069
25.7
NSW
Essential
Energy
Government
1,301,626
190,531
4,595
2,292
6.8
QLD
Energex
Government
1,316,295
53,928
8,120
4,875
24.4
QLD
Ergon
Government
689,277
160,998
7,380
2,429
4.3
TAS
Aurora
Government
275,536
25,844
1,410
1,760
10.7
SA
SA Power
Networks
Private
825,218
87,226
2,860
1,798
9.5
VIC
Citipower
Private
311,590
7,406
1,315
1,453
42.1
VIC
Jemena
Private
314,734
6,043
770
1,008
52.1
VIC
SPAusNet
Private
637,810
48,841
2,120
1,798
13.1
VIC
Powercor
Private
723,094
84,791
2,260
2,351
8.5
VIC
United
Energy
Private
641,130
12,875
1,410
1,962
49.8
It is notable that some of the DNSPs have particularly low customer densities, with the
lowest having an average customer density of only 4.3 customers per km of line (see Figure
3 for comparison). In rural areas, it is not uncommon for customers to be serviced by long
SWER feeders.
For this figure, eligible dwellings are defined as separate houses, semi-detached houses, caravans, cabins, and
improvised homes that may or may not be permanently occupied. It is sourced from the Australian Solar Portal
(forthcoming 2013) that uses data from the Australian Bureau of Statistics and the Clean Energy Regulator.
2
10
Particular distribution network hot spots for high PV penetrations have emerged driven by
a combination of factors including the housing stock, solar resource, local industry
capabilities, local government initiatives, and income levels. A number of isolated remote
grids have also seen significant household-driven PV system deployment, such as the towns
of Carnarvon and Alice Springs (case studies on these localities are presented in Section 0).
Figure 4: Percentage of eligible dwellings per post code with solar PV systems (status 03/2013)
11
A new subsection, Section 5A, of the NER contains provisions for interconnection for retail customers, including
embedded generators. See Upcoming Regulatory .
12
South Australian DNSP is not allowed to charge for any augmentation required as a result
of the connection.
The AER specifies network tariffs that can be charged by each DNSP in three-year price
determinations with annual revisions. These determinations are made on the basis of an
assessment of proposals submitted by DNSPs that operate in the NEM. In 2009, a DNSP
from New South Wales applied for permission to pass through administrative and
compliance costs associated with the connection of PV customers [21]. This request was
denied, and since this time no other similar requests have been made to the AER.
Medium-sized PV systems (from approximately 5 kW or 10 kW to 100 kW) have not been
eligible for feed-in tariffs in most jurisdictions in Australia, and larger system connections
must be individually negotiated with the DNSP. Fees are generally charged for processing
the application. Some DNSPs have a standard fee structure, whereas others price the
processing of applications individually [22]. Approval to connect is generally subject to a
network impact assessment, with additional associated fees and an uncertain time frame
(four to six weeks is not uncommon). The network impact assessment generally includes
both load-flow (static power flow) and protection studies carried out by the DNSP at the
expense of the PV customer. On the basis of the assessment, DNSPs may restrict export
from the PV system, offer the PV customer an alternative point of connection, or require the
PV customer to pay for network extension or augmentation.
The extent to which DNSPs can pass on any deep augmentation costs (network costs
associated with the connection that lie beyond the immediate network connection itself)
varies by jurisdiction. In Victoria, for instance, deep augmentation costs may not be passed
on, where shallow augmentation is defined as augmentation up to and including the first
transformer upstream from the generator [23]. DNSPs may also require additional
protection (above what is required by AS4777, the Australian standard for grid connection
of energy systems via inverters) and monitoring and/or control equipment and associated
systems. The size threshold above which systems are not automatically connected, the
process and timeline of assessment, and the fees differ by DNSP, and there is currently
limited transparency about the process or fees that may apply in many cases. In addition,
information about network capacity and existing loads in different parts of the distribution
network is not generally available in the public domain. A number of DNSPs are currently
preparing guidelines for the connection of medium-sized systems that will increase
transparency and certainty and streamline the connection process. A change to the
National Electricity Rules to increase the availability of distribution network data, including
zone-substation load data, is also currently under consideration.
AS4777 standards do not apply to large PV systems (>100 kW). Large systems are
invariably subject to individually negotiated connection agreements and usually require
more detailed technical studies than medium-sized systems. Systems up to 30 MW are not
required to formally participate in NEM scheduling processes, including making wholesale
market offers. In such cases, there appears to be negative load from a system operation and
planning perspective. Systems of less than 5 MW are also exempt from having to register as
generators.
13
AS4777
The most current version of the AS4777 standard is from 2005. The standard is currently
being revised, with a public consultation period completed at the end of August 2013. The
existing AS4777 standard is divided into three parts: installation requirements, inverter
requirements, and grid protection.
The key inverter requirements include
Power factor is to be in the range of 0.8 leading to 0.95 lagging for all output from
20% to 100% of rated output;
Total harmonic distortion (THD) (to the 50th harmonic) shall be less than 5%;
The inverter shall conform to the voltage fluctuation and flicker limits as per
AS61000.3.3 for equipment rated less than or equal to 16 A per phase and
AS61000.3.5 for equipment rated greater than 16 A per phase;
The inverter shall withstand a standard lightning impulse of 0.5 J, 5 kV with a 1.2/50
waveform; and
The direct-current (DC) output of the inverter at the AC terminals shall not exceed
0.5% of its rated output current or 5 mA, whichever is greater.
The grid protection device shall operate if supply from the grid is disrupted, when
grid voltage or frequency goes outside preset parameters, or to prevent islanding;
The inverter set points should be in the range of fmin = 45 Hz 50 Hz, fmax = 50 Hz 55
Hz, Vmin = 200 V 230 V, and Vmax = 230 V 270 V for a single-phase system;
The grid protection device shall incorporate at least one method of active antiislanding protection; and
Reconnection is permitted when voltage and frequency are in the acceptable range
for at least 1 minute and the inverter energy system and the electricity distribution
network are synchronized and in phase with each other.
In the proposed revision of AS4777, parts two and three of the current standard have been
combined. The key changes are listed in Section 0.
15
AS61000.3.100
The AS61000.3.100 standard, published in 2011, sets a nominal supply voltage of 230 V,
although the LV system was originally designed and specified at 240 V. It defines both an
allowable operating range and a preferred operating range. The preferred operating range
is +6% to -2%; however, the allowable voltage variation at the point of supply is +10% and
-6%. The preferred operating range represents the 50 percentile value of voltage, whereas
the upper and lower limits are the 99 and 1 percentile values, respectively (see Table 5 and
Figure 6).
Table 5: System Voltage Requirements Under AS6100.3.100
Voltage
Max. allowable
Max. preferred
Nominal
Min. preferred
Min. allowable
AS61000.3.100
253 V (99%)
244 V (50%)
230 V
225 V (50%)
216 V (1%)
AS3000
Section 3.6.2 of the AS3000 standard specifies that a voltage drop of no more than 5% is
allowed between the point of supply and any point in the electrical installation. The voltage
drop limit applies between a solar inverter and the point of supply [24].
16
Limit
ACT
Voltage
High
254
Nominal 240
Low
226
Frequency High
50.25
Nominal 50
Low
49.75
New
South
Wales
253
230
216
50.25
50
49.75
254
240
226
50.25
50
49.75
253
230
216
50.25
50
49.75
253
230
216
50.5
50
49.5
253
230
216
50.25
50
49.75
254
240
226
51.25
50
48.75
DNSP Regulations
The DNSPs develop regulations based on the requirements of the state and territory
regulator and the Australian Standards. These regulations address the connection
procedure and the technical guidelines with which a PV system must comply. DNSPs
typically set a system size limit above which a more detailed technical study is required.
This is to provide a streamlined connection process for household customers. Some DNSPs
also specify the voltage and frequency set points of the inverter. The set point values,
where available, and the system size limits are summarized in Table 7.
17
Table 7: System Size Limits and Voltage and Frequency Inverter Set Points
State
DNSP
f(max)
V(min)
V(max)
48 50
48
48 50
46
50 52
52
50 52
54
200
190
200
210
260
260
260
270
ACT
ActewAGL
NSW
NSW
NSW
NT
AusGrid
Endeavour
Essential Energy
Power and
Water
Corporation
QLD
QLD
Energex
Ergon
5 kW
5 kW, 2 kW for SWER
48
-
52
-
255
255
TAS
Aurora
SA
48
52
257
VIC
SA Power
Networks
CitiPower
48.5
51.5
195
265
VIC
VIC
Jemena
SPAusNet
10 kW, or 30 kW three
phase
10 kW, 5 kW for SWER
network
10 kW per phase, unless
rural location
10 kW per phase
3.5 kW per phase SWER, 4.6
kW single phase, 5 kW per
phase, three phase
VIC
VIC
WA
Powercor
United Energy
Western Power
10 kW per site
10 kVA per phase
5 kVA single phase, 30 kVA
three phase
47.5
52
WA
Horizon Power
10 kW per phase
46.5
53
190
265
Some DNSPs have introduced, or are considering introducing, capacity limits on the
installed PV that can be connected to any one feeder, substation, or other network element.
For example, Western Power has limits on embedded generation of 30% of the LV feeder
capacity and 20% of the zone substation capacity [27].
18
Another example is from ActewAGL, which states that systems greater than 61 kW will be
considered for SCADA integration with the network operator [29]. SA Power Networks
specifies additional islanding protection of a suitable protection relay with both Vector
Shift and ROCOF protection elements incorporated with the relay [30]. The system must
also record and transmit to the utility quality-of-supply data, including THD and long- and
short-term voltage flicker measurements. This is for systems of more than 10 kVA per
phase.
19
21
Figure 8: Diagram of PV system distribution on the Alice Springs Network (status: 10/2010)
22
Comment/Status
2.1.1.1.1
Previously experienced during certain system lowfrequency events. Steps have been taken by P&W
to address this by changing inverter low-frequency
trip requirements (i.e., reduced to 46 Hz). This issue
has been resolved for the connection of future PV
systems but not yet fully resolved for existing PV
systems on the network. There has been no
significant impact on network operation.
2.1.1.1.2
2.1.1.1.3
2.1.1.1.4
PV system islanding
Harmonic injection
23
24
System-Wide or
Localized
Current/Proposed Management
Strategies
PV systems
impact on
network
stability due
to inverter
anti-islanding
protection
detecting
significant
frequency
deviations
System
Current:
Trial:
Proposed:
25
Current:
Rectification of phase
imbalance with respect to
both loads and PV system
connections
Load shifting
Network augmentation
Trial:
PV system
impacts on
network
stability due
to cloud
fluctuations
System
Voltage regulation
technology
Current:
Trials:
Proposed:
26
Further monitoring of
system loads and PV
generation
Fires due to
PV systems
Localized
PV system
impact on
planning
strategies
System and
localized
Change to problematic
junction box designs
Current:
Trial:
System
islanding
System and
localized
Current:
Proposed:
27
PV inverter protection
settings are being reviewed
in line with the impact on
system stability and in line
with studies mentioned
above. Horizon Power would
prefer that all inverters are
set to a fixed value rather
than be variable inside a
range.
System
harmonics
from PV
inverters
Localized
Proposed:
Reverse
power flow
Localized
Proposed:
Reduction in
System
generator fuel
use
Benefit:
Offsetting of
peak summer
loads with PV
generation
PV generation generally
corresponds well to the peak
system loads, implying possible
deferral of network upgrades,
and benefits can be further
maximized by adjusting
customer loads.
Benefit:
System
28
Monitoring at higher PV
system penetrations is
important to ensure that the
PV
network harmonics into the
future
Monitoring at higher PV
system penetrations and a
review of protection
schemes is needed to
prevent potential future
problems
There is potentially
significant value in such fuel
savings depending on
gas/diesel prices. The value
of climate change
abatement with PV is also
potentially significant. By
managing the spinning
reserve strategy effectively
and increasing the amount
of PV in the system, these
benefits can be maximized.
PV Penetration
PV
Measure
Measure
Installed nominal
PV capacity
PV capacity
penetration
Est. summer
PV peak power
midday
PV peak
penetration
power
summer
PV peak power Est. winter midday
PV peak power
penetration
winter
Est. average
PV peak power
midday PV peak
penetration
power
average
Est. annual PV
PV annual
energy
energy
penetration
PV generation at
Maximum
instantaneous PV time of max. PV
penetration
penetration4
Estimated
Value
1,102 kW
System
Measure
Annual peak
load
Average
summer
midday load
Average
winter midday
load
Average
midday load
% PV
Penetration
5,050 kW
22%
2 GWh
698 kW
583 kW
497 kW
392 kW
Value
2,914 kW
20%
1,984 kW
25%
2,372 kW
16.5%
Annual gross
system load
39 GWh
5%
Load at time
of max. PV
generation
2,158
kVA
32%
30
for this report. The main issue experienced by Ergon Energy due to high penetrations of PV
has been voltage related. None of the other potential power quality issues such as reverse
power flow, power factor distortion, or harmonics are currently of significant concern.
Protection staff, however, raised network stability concerns in relation to islanding, though
there is no evidence of this having occurred to date. All instances of voltage issues that have
arisen have been addressed by Ergon Energy, and the company continues to trial and,
when appropriate, introduce new technologies and procedures for both mitigating current
voltage issues and avoiding new problems as PV capacity on Ergon Energy s network
increases.
Table 11: Summary of the Experiences of High PV Penetration in Magnetic Island and Townsville
Current:
General:
Trials:
31
LV regulators
Phase Imbalance
Current:
Current:
Current:
Current:
33
Harmonics
Current:
Proposed:
34
Energy is proactively exploring management approaches for dealing with the higher PV
penetrations, and other distributed energy technologies, that seem certain to come.
36
The voltage drop between the inverter and the main switchboard is to be less than
1%, and the voltage drop between the main switchboard and the distributor point of
supply is to be less than 1%.
An inverter must have the ability to offer power quality response modes through
volt-watt response, volt/Var response, reactive power support, and power ramprate limits.
An inverter will support demand response modes (DRM) 0 through to 5, and modes
6 to 8 should also be supported (refer to Table 12).
Table 12: Demand Response Modes for Inverter Performance
Demand
Response Mode
DRM0
Requirement
Operate the disconnection device
DRM1
DRM2
DRM3
DRM5
DRM6
DRM7
DRM4
DRM8
The demand response capabilities that are anticipated in the revised version of AS4777 will
be complemented by a new section of the AS4577 standard. This standard contains
provisions for the control of electrical devices that can alter electricity demand in response
to an instruction from a remote agent. In practice, this remote agent is the DNSP.
AS4577 already provides operational instructions for the use of demand response
functionality in air-conditioners and swimming pool pumps. PV inverters will be covered in
an upcoming addition.
37
38
Austria
Authors:
Benoit Bletterie,1 Roland Brndlinger,1 and Christoph Mayr1
1Austrian
39
Value
0.61 GW
11.3 GW
23.2 GW
337.5 GWh
0.5%
0%
0%
5%
No data
95%
No data
Year
Status: 12/2013 [35]
REF2_2_1 and PV Austria
estimation
Status: 12/2012 [35]
Status: 12/2012 [35]
Status: 12/2012 [35]
Status: 12/2012 (estimation)
Status: 12/2012 (estimation)
Status: 12/2012 (estimation)
-
Figure 11: Development of PV capacity in Austria from 2000 to 2013 (Source: PV Austria
http://www.pvaustria.at/2013/11/pv/)
40
41
Requirement
Reference
[40]
42
The reactive power control strategy and set points are defined by the DNO and depend on
local grid conditions. For this purpose, the TOR D4 defines a number of options that may be
necessary, such as
For generation systems connected to the MV level, the provision of reactive power can be
either realized autonomously (e.g., fixed preset power factor or voltage-dependent reactive
power provision Q(U)) or via remote control. For generation connected to an LV grid, only
autonomous (without communication link to DNO) reactive power provision methods are
used.
Active Power Control:
Generation plants with a system capacity of more than 100 kW need to implement active
power control on demand of the DNO.
For smaller plants, there are no common requirements for active power control; however,
some DNOs require an external control unit for smaller plants above 5 kW.
43
KK2
PV5
3,06 kW
KK3
PV4
1,7 kW
PV1
2,38 kW
PV3
5,13 kW
KK1
PV2
3,06 kW
To investigate the behavior of the PV inverters with local voltage control features, several
measurement systems have been used. The measurement data have been stored in a
common database for ease of further processing.
-
In addition to the functional validation (to demonstrate that the voltage control
implemented into the inverters behaves under real conditions as intended), another
44
objective of the field tests was to quantify the benefit of this control and compare it to the
values obtained in simulations.
To quantify the compensation of the voltage rise achieved with the different control modes,
a validation concept has been developed. It consists of comparing the cumulated
probability distribution of the voltage for a control mode and for a reference case (without
control). To limit the impact of external effects (voltage variations in the MV network, nonconstant load conditions), only sunny working days have been considered (see [46] for a
detailed description of the concept). In addition, instead of directly using the voltage
values, the voltage difference between a distribution cabinet and the last PV installation has
been used. This allowed eliminating the impact of the MV network in the validation
process. Further, the investigated control modes (Q(U), Q(U)&P(U), cos fix, and
cos(P)) have been parameterized on the basis of the measurements from the reference
phase (without control). The voltage levels are not close to the absolute limit (because the
voltage level of the MV network is rather low in this area), thus the control settings have
been moved to smaller voltage values. The settings of the Q(U) and P(U) controls
mentioned in the following are:
Q(U):
U0,98 p.u. Q=Qmax=Snsinmax=0,43Sn;
U<1.01 p.u. und U>0.99 p.u. Q=0;
U>1.03 p.u. Q=-Qmax=-Snsinmax=-0,43Sn;
P(U)
U>1.03 p.u. P =P(U) curtailment with 100%/V
Figure 13 and Figure 14 show some of the results of the functional validation for a
Q(U)&P(U) control in the form of the PQ and Q(U) diagrams, respectively. The basis for
these diagrams is the high-resolution power quality measurement data at the farthest PV
installation. The reduction of the reactive power for voltage values greater than 2.5%
shown in Figure 14 is due to the power factor limitation (cos>0.85) combined with the
P(U) control.
45
Figure 13: PQ-Diagram of Inverter 3 3-s RMS average [Test Q(U)&P(U)] [46]
Figure 14: Q(U) diagram of Inverter 3 3-s RMS average [Test Q(U)&P(U)] [46]
46
In addition to the data analysis performed for the functional validation, the compensation
of the voltage rise due to the voltage control modes has been evaluated. Figure 15 shows an
example of the evaluation of the Q&P(U) control compared to the reference case (without
control). The effect of the control can be mainly observed as expected for HV values. The
compensation of the voltage rise between these two distribution cabinets amounts to about
0.75%, which represents about 2.25% of the nominal voltage when considering the whole
feeder.
Figure 15: Cumulated distribution of the voltage difference between two distribution cabinets for
sunny working days 10-min RMS average values [Test Q(U)&P(U)] [46]
47
2013-08-16 14:00-16:00
1
0.9
0.8
P/P
stc
0.7
0.6
0.5
0.4
0.3
0.2
14:00:00
ID 1
14:30:00
ID 2
15:00:00
hour of day
ID 3
ID 5
15:30:00
Irradiation
16:00:00
ID 3- Highres
Figure 16: Active power curtailment from the P(U) control 3-s and 5-min RMS average
[Test Q(U)&P(U)] [46]
Figure 16 shows the normalized (STC) active power infeed of four of the PV installations
(the fifth one is not controlled) between 14:00 and 16:00 on a sunny day. This figure shows
that only the two farthest installations actually curtail their output power (about 2.5% of
the yield for this period).
As a conclusion of these field tests, the effectiveness of the control modes could be
validated: the voltage rise could be compensated by 1.5% (reduction of almost 30%) with
reactive power only, which corresponds to the value obtained by the simulations.
In addition to these investigations, numerous simulations have been performed to try to
generalize the conclusions. A few of the findings are summarized below:
(1) The compensation of the voltage rise in cable network is significantly smaller than
in overhead line networks; however, a compensation of almost 20% can be reached
with common cables (e.g., 150 mm2 see equation (1) and Table 15).
(2) The voltage drop over the transformer reactance can be rather large for old, small
transformers (large, short-circuit voltage and low power) 1.5% of the nominal
voltage for a 50% penetration and a 6% short-circuit voltage. See equation (2) and
Table 16.
(3) As generally known, reactive power control is more effective in overhead line
networks. In addition, a small transformer with a large, short-circuit voltage leads to
an even stronger effectiveness. Such conditions are generally met in older networks.
48
However, considering that such networks are generally upgraded partly for other
reasons than high PV penetration (e.g., reliability), the actual value of reactive
power based voltage control in older networks must be carefully considered.
(1)
Table 15: Compensation of the Voltage Rise for Cable and Overhead Lines (for cos=0.90)
Overhead Line
R/X
Compensation
@cos=0.90 (%)
1.9
26.1
1.4
34.4
1.1
45.6
0.8
57.1
Cross-Section
(mm2)
50
70
95
120
150
240
Cable
R/X
7.2
5.3
3.8
3.2
2.6
1.7
Compensation
@cos=0.90 (%)
6.8
9.1
12.9
15.2
18.6
29.2
(2)
Material: Aluminium
49
Table 16: Voltage Drop Over the Transformer Main Reactance (for cos=0.90)
P/SN=10%
0.2
0.3
uk=4%
uk=6%
P/SN=20%
0.4
0.6
P/SN=50%
1
1.5
P/SN=100%
1.9
2.9
Finally, a short economic analysis of active power curtailment has been performed within
this project [44]. In this study, the same feeder has been considered, but the PV penetration
has been highly increased (3 kWp per house).
With this high PV penetration scenario, some installations must curtail their active power
output. The annual yield loss has been assessed and is shown on Figure 17. Although the
maximal curtailment reaches more than 7% of the annual yield for a single installation, the
overall yield loss for the whole feeder amounts to only 0.8%.
1 2 3 4 5 6 7 8 9 101112131415161718192021222324252627282930
PV-installation
Financial impact per year
In addition, the financial impact of this curtailment has been evaluated for prosumers with
a self-consumption regime and for a feed-in-tariff regime. For this the exact conditions
must be considered because the value of the amount of lost energy depends on whether or
not it would have been self-consumed (Figure 18). The maximal curtailment of almost 7%
of the annual yield represents about 23 year, which corresponds to about 4% of the
revenues that this installation would generate with the current FIT.
50
7000
6000
5000
4000
3000
2000
1000
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
hour of day
Figure 18: Self-consumption diagram for Installation 23 (on a sunny summer Sunday) [44]
1 2 3 4 5 6 7 8 9 101112131415161718192021222324252627282930
PV-installation
Financial impact per year
/year
-5
-10
-15
-20
-25
1 2 3 4 5 6 7 8 9 101112131415161718192021222324252627282930
PV-installation
51
IRR (%)
0
0
10
15
PV-installation
20
25
30
Figure 20: Impact of the power curtailment on the IRR for the 30 installations [44]
Finally, the impact of the power curtailment on the profitability of the PV installation has
been quantified by analyzing the internal rate of return (IRR) for a lifetime of 25 years. For
this a total cost of 6,750 per installation ( kWp 2,250) and a subsidy of 400 kWp have
been considered. The results of this analysis are shown in Figure 20, where the red area
corresponds to a decrease of the IRR due to power curtailment. Depending on the rate of
self-consumption, the internal rate of return is between 0.5% and 6% per year. This figure
shows that for the most affected installation (23), the maximal financial impact is about
0.7 percentage points (IRR decreases from 3.5% to 2.8%).
In conclusion, voltage control is a valuable alternative to network reinforcement to
enhance the hosting capacity of LV networks. In addition to the Q(U) control, which allows
compensating part of the voltage rise caused by the PV infeed, a P(U) control has been
implemented to guarantee that the voltage limits can be met. The P(U) control is seldom
activated but ensures that the voltage limits are met while avoiding repeated
disconnections from the grid.
The simulations made for a high penetration case study assuming 3 kWp/roof showed that
the amount of active power curtailment necessary to maintain the voltage below the limit
is rather limited (up to 7% for the most affected installation and less than 1% for the whole
feeder). The resulting financial impact is also limited and is smaller under a selfconsumption regime than under a feed-in-tariff regime. However, the regulatory
framework for such an active power curtailment at the distribution level is still missing and
shall be addressed to allow its deployment. By implementing such a voltage control, the
hosting capacity could be extended by more than 40% compared to the case without
52
control when considering a smart planning approach and by more than 800% when
considering existing planning rules. These figures shall be carefully used because new
planning rules are needed for networks with a high share of generation featuring smart
grid functionalities. The relation between self-consumption rate and revenue losses has
been investigated. The actual impact of the P(U) control depends on the network topology
and the distribution of the installations (load and PV) along the feeder and among the
phases. The impact of the individual local situation plays only a minor role compared to the
overall power-flow: ensuring a high self-consumption at one particular installation does
not necessarily means that the impact on the network (e.g., voltage rise) is low and that
curtailment can be avoided.
The behavior of the neighbor prosumers is influencing the voltage, and a dedicated shifting
of the load (with or without storage) has a positive influence, especially if the whole
neighborhood is participating. Due to local effects and a potential high imbalance between
phases, the use of a single-phase storage system to increase the self-consumption can
worsen the voltage profile.
Lack of standardization,
In particular, the use of a reactive power based voltage control alone does not provide any
guarantee that the voltage will stay within the limits. Of course, over-voltage situations are
in practice impossible due to the over-voltage protection, but when seeking a smart way of
meeting the voltage quality targets while maximizing the PV penetration, proper solutions
are needed. In addition to a reactive power based voltage control, an active power control
can provide the previously mentioned guarantee. However, many regulatory questions
must be answered beforehand (e.g., allocation/repartition of the curtailment costs).
Another topical issue is the billing of reactive power from generators participating or not
participating in the system operation (direct reactive power provision or reactive power
based voltage control).
53
Belgium
Authors:
Karel De Brabandere1 and Carlos Dierckxsens1
13E,
Brussels, Belgium
54
Value
2,501 MW
14,234 MW
20,813 MW
74,985 TWh
84,758 TWh
1,628 TWh
2%
<<
<
>>
9 kW
Year
Source:[47], Section 2.3
Source: [47], Table 3.3.2.2.1
Source: [47], Section 2.3
Source: [47], Section 2.2
Source: [47], Section 2.2
Source: [47], Section 2.2
Source: [47], Section 2.2
Status:
Status:
Status:
Status:
55
voltage problems cannot be directly solved. In that case, the DSO will need to reinforce the
grid or use storage, curtailment, or reactive power control, as will be discussed below.
MVA
MVA
MVA
57
Islanding:
Islanding within a private installation is allowed. Islanding situations in which the public
distribution grid is involved are prohibited.
Fault ride-through (FRT):
PV installations are not allowed to disconnect during voltage dips reaching voltage levels
below the line indicated in Figure 22.
100
90
Residual Voltage (%)
80
70
60
50
40
30
20
10
0
0
0.5
1.5
2.5
Duration [s]
58
PF = f(active power P)
Q = f(voltage U)
Figure
24:23:
ForFor
all but
oneone
inverter,
thethe
reactive
power
controls
reduce
thethe
probability
of voltages
Figure
all but
inverter,
reactive
power
controls
reduce
probability
of voltages
exceeding
250
V
(as
far
as
they
are
not
zero).
For
that
one
inverter,
the
probability
of
exceeding
exceeding 250 V (as far as they are not zero). For that one inverter, the probability of exceeding 251
251 VV
was
wasfound
foundlower
lowerfor
forQ(U)-b
Q(U)-bthan
thanwithout
withoutcontrol.
control.Circles
Circleson
onthe
thesame
samevertical
verticalline
linebelong
belongto
tothe
thesame
same
inverter.
The
size
of
the
circles
is
a
measure
for
the
available
measurement
data.
For
the
circles
inverter. The size of the circles is a measure of the available measurement data. For the circles at
at the
the
far right,
there
waswas
not not
sufficient
datadata
without
reactive
power
control
within
the the
measurement
far right,
there
sufficient
without
reactive
power
control
within
measurement
period.
period.
59
Economic Evaluation
A methodology for comparing different implementation options is described in [52]. The
concrete outcome of these cost analyses are subject to many assumptions on the eventual
implementation of the considered solutions. Important aspects that need to be taken into
account are, among others:
The voltage level on which the PV is connected, the related costs for network
modifications, and the electrical characteristics, reflecting the performance of different
solutions (e.g., a higher electrical reactance of the network feeders favors reactive
power control);
The foreseen required hosting capacity in the future. Are solutions implemented in a
stepwise approach or rather imposed for all as a retroactive change in the requirements
for existing systems?;
For distributed solutions such as local reactive power control, whether and how
installations with different properties (local voltage level, reactive power capability of
the inverter, system size) are treated needs to be considered;
The distribution of the loads and generators between the phases of the feeder; and
The economic analysis is based on data of one complete year and a uniform electric load
over the feeder. In the simulation tool, it was assumed that all OPEX solutions, which do
not require any additional equipment, are exhausted. This means that the (fixed)
MV/LV transformer tap setting is already optimized to lower the LV voltage as much as
possible without causing under-voltage problems during high load conditions. Also, it is
assumed that all generation and loads are evenly distributed over the three phases to
obtain balanced voltages.
More technically, all the assumptions applicable in [52] were followed, with the additional
assumptions that:
All inverters on the feeder are retrofitted to have reactive power capabilities;
The central (supervisory/ coordinated) reactive power control is able to gather the
feeder voltages in real time and optimally dispatch the inverters;
The local reactive power control is implemented as a Q(U) function that is able to keep
the voltage below the limit at all times. Note that with this control the injection of
reactive power will no longer be the same for all inverters on the feeder;
Storage costs amount to 500 EUR/kW for the converter and 364 EUR/kWh for the full
setup behind the converter (lead-acid technology); and
60
o Real-time control over private SCADA network that allows the DNO to change
the inverter active and reactive power settings in real-time and in a secure
way;
o Standard control over private SCADA network that allows the DNO to change
the inverter settings/characteristics in a secure way but with a certain delay;
o Standard control over the internet that allows the DNO to change the inverter
settings/characteristics in a certain delay such that security is limited to
access authorization requirements of inverter; and
o Standard control over the internet (inverter integrated communication) that
is the same as above but with reduced infrastructure costs thanks to inverter
integration.
Discussion of Results
The results allow comparing grid reinforcement to inverter control in different aspects,
including:
Different levels of hosting capacity increase;
Local (Q(U), P(U)) versus central (supervisory) inverter control;
Different implementations of local inverter communication;
Curtailment versus storage; and
Low versus MV feeders.
All these aspects are brought together in one overview for LV (Fehler! Verweisquelle
konnte nicht gefunden werden.) and one for MV (Figure ). The x-axis shows the increase
in hosting capacity, and the y-axis describes the corresponding cost. The bars show the
costs without communication equipment for each scenario. The markers above the bars
indicate the total cost including the communication equipment. Note that central control
always requires real-time control equipment. Also note that the cost of network
reinforcement is approximately constant because these works are typically not performed
in an incremental way. Moreover, the civil works largely outweigh the cost differences
between the different cables required for each case.
As a general observation, energy storage is relatively expensive compared to both energy
curtailment strategies (yellow versus red, green versus blue) and grid reinforcement
(yellow, red versus grey). Although storage is under continuous development and different
technologies could be used, the current price limits the economical use of storage to cases
in which very limited amounts of storage can avoid more expensive grid reinforcements.
This restriction is more pronounced on the LV feeders.
61
On LV, in all cases where inverter control is more efficient than grid reinforcement, local
control strategies come out as the most economically viable. However, central control is
overall more efficient in terms of reactive and active power efficiency (OPEX), and these
benefits do not outweigh the additional communication needs (CAPEX) in this case.
One of the project objectives was to allow DNOs to increase the grid s hosting capacity for
PV by 50% at only 10% of the costs of traditional grid reinforcement. The results show that
for LV feeders, this would be feasible with active power curtailment and using inverter
integrated modules for communication. These are two difficult constraints because (1)
current European Union regulation does not foresee renewable energy curtailment at the
distribution level and (2) inverters with integrated communication are new (as are reactive
power capabilities) and existing PV systems can take advantage of this only when the
complete inverters need to be replaced. Note that although communication is not
absolutely required for local control, it may be preferable to prevent a lock-in with
unchangeable inverter settings at a later stage when the network requirements change.
Figure 24: Increasing the LV grid-hosting capacity for PV with inverter control functionalities can be an
economical solution for small increases in hosting capacity and basic communication infrastructure.
62
Figure 25: Inverter control can be a good way to increase the hosting capacity on MV levels. Due to
the higher impact on reactive power, storage (or curtailment) needs become important only above a
50% hosting capacity increase.
To increase the hosting capacity with 50% on MV feeders, all solutions are more costefficient than grid reinforcement. This is attributed to the high impact of reactive power
(low R/X ratio of the grid) and the relatively larger PV system size, reducing the
communication cost per installed PV capacity. Local reactive power and storage control is
the most efficient and stays below 15% of the grid reinforcement cost. Note that in this
particular case storage performs better than curtailment because the network losses are
being reduced during the storage discharge period.
63
China
Author:
Wang Yibo1
1Chinese
64
Value
8 GW
4.5 GW
674 GW
1,145 GW
3,700 GWh
0.1%
63.6%
36.4%
Year
Status: 2012 [53]
Status: 2012 [53]
Status: 2012 [54]
Status: 2012 [55]
Status: 2012 [56]
Status: :2012 [56]
Status: 2012 [53]
Status: 2012 [53]
The HV distribution network is usually supplied by 35 kV, 66 kV, or 110 kV, as well as
220 kV for high-density load.
HV Level: During the normal operating condition, the voltage bias of the point of common
coupling (PCC) between a PV plant and power system should between -10%UN ~+10%UN.
Five typical structures are commonly used for the HV network in Chinese city, including
radiation, ring, link, T-type, and hybrid wiring. The radiation type is suitable for a 110(66)kV end substation and a 110(66)-kV station without a second source The -wiring with
two sources is applicable in the areas where there are needs for strong connection,
frequent changes in operation, or severe requirements for load density and reliability
between two 220-kV-substations. The chain-wiring with two sources applies in situations
when
1) two or three 110(66)-kV substations are located in the same transmission corridor,
which can supply source lines on the unified tower; or
2) there are needs for strong connection, frequent changes in operation, or severe
requirement for load density and reliability between two 220-kVsubstations.
The hybrid wiring is suitable for regions in which load grows very fast and requires high
reliability.
65
Capacity
Grid Layout
Voltage
System topology
Phase configuration
Installation type
Overhead lines/cables
Typical nominal capacity of DG
G 6 MWp
MV Level: The MV distribution network in China is mainly based on 10 kV. Some large
industrial enterprises use 6 kV. During the normal operating condition, the voltage bias of
PCC should be between -7%UN~+7%UN. The MV distribution network includes two basic
wirings: the radiation-type wiring and the ring-type wiring, by cable and overhead line.
10 kv/6 kV
7% UN,LL [57]
System topology
Phase configuration
Three-wire/three-phase
Installation type
Overhead lines/cables
G 6 MWp
Capacity
Grid Layout
Voltage
66
220 V/380 V
System topology
Phase configuration
Three-wire/three-phase
Installation type
Overhead lines/cables
NA
Capacity
Grid Layout
Voltage
strengthen the effect of national plans on local plans, establish a protective regulation of
totally purchasing RE electricity, and establish an RE development fund.
Technical Framework
It is the responsibility of the supply side to ensure the power quality, including technical
parts (i.e., voltage quality) and non-technical parts (i.e., service quality). The power quality
is essential for safe production and fruitful profit of the industry as well as people s living.
After a PV system is connected to the grid, the voltage bias of the PCC should meet the
requirements of GB/T 12325 [59]. The voltage fluctuations and the flicker should meet the
requirements of GB/T 12326 [60]. Table 22 shows the voltage fluctuation requirements
after PV generation system connect to the grid.
Table 22: Requirements for Voltage Deviation in China s Distribution Grids According to GB/T 12326
d/%
r/( times/h)
LV/MV
HV
r 1
1<r 10
3*
2.5*
10<r 100
1.5
100<r 1000
1.25
Note:
1) The r represents the frequency of voltage change, which refers to the count of the voltage change
per unit time (ascending or descending voltage changes counted once independently). Changes in
several different directions, if the interval is less than 30 ms, should be counted once. The d
represents the voltage changes, and its value is the voltage difference between the two adjacent
extremes of the rms curve, expressed by nominal voltage of the system as a percentage.
2) For small changes in the frequency of r (less than once per day), the voltage change limits can be
relaxed.
3) For random irregular voltage fluctuations, the mark "*" is labeled for its limits in the table.
68
69
Q(pu)
0.31
P(pu)
-0.31
When the reactive power rating of the inverter does not meet the need of voltage
regulation, it should install an appropriate reactive power compensation device in the PV
power station. If necessary, a dynamic reactive power compensation device should be
installed.
A PV power station connected at the 10-kV to 35-kV level should have the ability to adjust
reactive power and participating voltage regulation according to the voltage of the PCC.
The adjustment method, the reference voltage, and the voltage transfer slip should be set
by the scheduling institute.
For a PV power station connected at 110 kV or above, a reactive power and voltage control
system should be equipped. According to the instruction of the scheduling institute, a PV
power station should automatically adjust its reactive power output or input and perform
voltage control of PCC, and the adjusting speed and control precision should meet the
requirements of the voltage regulation of the power system. When the utility grid s voltage
is within the normal range, a PV power station connected to a 110-kV-level should be able
to control the voltage of the PCC in the range of 97% to 107% of nominal voltage. A PV
power station connected to a 220-kV-level or above should be able to control the voltage of
the PCC in the range of 100% to 110% of nominal voltage.
70
distributed PV projects, including the high-tech industrial park Xiuzhou distributed solar
PV power generation project phase (61 MW), the Tongxiang Economic Development Zone
of distributed PV demonstration phase (15 MW), and the Haining tech zone distributed PV
power generation project (20 MW). With the changes in China s policies, market-oriented
distributed PV systems will become the mainstream of the PV market in 2015, and by 2020
the share of the distributed PV market will exceed 50%.
71
72
Germany
Authors:
Thomas Stetz,1 Markus Kraiczy,1 and Konrad Diwold1
1Fraunhofer
73
Value
36.2 GW
80 GW
132.7 GW
28.000 GWh
5.7%
.6%
28.8%
61.7%
W
Year
Status: 03/2014 [64]
Status: 2011 [65]
Status: 2011 [65]
Status: 2012 [66]
Status: 2012 [67]
Status: 06/2013 [68]
Status: 06/2013 [68]
Status: 06/2013 [68]
Status: 06/2013 [68]
Status: 06/2013 [68]
Status: 2012
Figure 27: Installed PV capacity per ZIP-code area (status 05/2013). Figure by Y.M. Saint-Drenan,
Fraunhofer IWES
74
monitoring of overhead lines, and load monitoring with automatic n-1 calculations to fully
utilizing the reserve capacities of conductors during normal operation.6
Controllable Equipment:
+ Remote-controlled SVC and STATCOMS for reactive power compensation
+ Autonomous and manual (remote) tap settings at the substation transformer
+ Automatic and manual (remote) grid topology reconfigurations
+ Reactive power control of DG within specified limitations, according to [74] (see also 0)
+ Active power control of DG during disturbed grid operation, according to [74] and the
German EEG.
Major Technical Barriers for PV Interconnection: Major technical barriers for PV
interconnection at the HV level are typically transmission capacities at the HV level and
substations. However, currently only 5.6% of the total installed PV capacity is connected
directly to the HV level (in total about 1.8 GWp).
Table 24: Basic Information on HV Level
Load and
Generation
Grid Layout
Voltage
110 kV
100 kV/123 kV
96 kV/127 kV
System topology
Mostly meshed
Phase configuration
Three-wire/three-phase
Installation type
Overhead lines/cables
Multi-MW PV system7
Typical customer
See E.ON Netz demonstration line segments Flensburg-Niebll and Flensburg-Breklum. Reference:
http://www.forum-netzintegration.de/uploads/media/20120918_Jansen_KreisNF.pdf
7
E.g., PV plant Perleberg, 35 MWp, Brandenburg
76
MV Level: The rated capacity of HV/MV transformers is usually between 25 MVA and 63
MVA. The preferred nominal line-to-line voltage at the MV level is 20 kV 10%
(requirements defined by EN50160). In suburban areas also, 10-kV systems can be found.
Some few supply grids are operating with a nominal voltage of up to 35 kV. The MV level is
realized as a three-phase delta system consisting mostly of overhead lines in rural areas
and underground cables in suburban areas. In cases of grid reinforcement and/or
expansions, underground cables are preferred. Meshed system topologies can be found in
urban areas. In suburban and rural areas, often a mixture of open and closed-loop
structures exist. In some rural areas, pure branch-feeder configurations can be found.
Typical interconnection customers are large industrial facilities as well as utility-scale
generation units with a rated capacity of up to some MW.
Power System Management: DSOs, responsible for MV system operation, are authorized
to switch circuit breakers and disconnectors at the MV level only (remote control). SCADA
systems monitor voltages and power flows at selected crucial MV nodes (e.g., transformer
substations, switching substations). Current technical trends are the incorporation of MV
state-estimation for monitoring purposes and expert systems for sending active and
reactive power set values to PV plant controllers.
Controllable Equipment:
+ Autonomous and manual (remote) tap settings of the transformation ratio at the
substation transformer
+ Manual (remote) grid topology reconfigurations
+ Reactive power control of DG within specified limits, according to [74] (see also 0)
+ Active power control of DG in disturbed grid operation, according to [74] and the German
Renewable Energy Sources Act EEG 2012 6, 11 (see also 0).
Major Technical Barriers for PV Interconnection: Currently, about
of Germany s
total PV capacity is directly connected to the MV grid and 1.1% to HV/MV substations (in
total about 10.4 GWp). Over-voltages and over-loadings of conductors and transformers
are the major technical barriers for PV interconnection, especially in rural areas.
Traditional grid reinforcements are the preferred solution to increase the hosting capacity
for PV at the MV level.
77
Load and
Generation
Grid Layout
Voltage
Basic Technical Guidelines: Distribution Code 2007 [73], Technical Guideline for the
Connection and Parallel Operation of Generators Connected to the MV Network [74]
Nominal line-to-line voltage
10 kV/20 kV/30 kV
Vmax 2% [74]
System topology
Phase configuration
Three-wire/three-phase
Installation type
Overhead lines/cables
1 km 20 km [85]
0.25 3 MW
Typical customer
78
LV Level: Delta-star transformers with a rated capacity of typically 100 kVA to 630 kVA are
used for the link between MV and LV levels. In grid sections with high local PV penetration,
transformer ratings of more than 630 kVA are also possible. The nominal line-to-line
voltage at LV levels is 400 V 10% (requirements defined by EN50160). The system is
realized as a three-phase star system with mostly underground cables comprising a neutral
conductor. In some rural areas, overhead lines still can be found. Most of the LV grids
consist of branch feeders only. For earthing TN-C-S systems, as defined by IEC 60364-1, are
most common. Typical interconnection customers are private households and small
commercial buildings. Most of the German PV capacity can be found at LV levels (70% of
the total PV capacity [68]).
Power System Management: Typically, no active monitoring and control are applied.
Current development trends are LV state-estimation and on-load tap changers for MV/LV
transformers.
79
Controllable Equipment:
+ Manual (not remote) topology reconfigurations at switching cabinets
+ Manual (not remote) transformer ratio settings of MV/LV transformers (usually not
equipped with on-load tap changers for voltage control)
+ Autonomous reactive power control of DG within specified limits, according to [75]
+ Remote active power control of DG during disturbed grid operation or a general active
power curtailment of small DG (<30 kWp) according to [75] and the German Renewable
Energy Sources Act EEG 2012 6, 11 (see also 0)
Major Technical Barriers for PV Interconnection: Over-voltages are the major technical
barrier for PV interconnection, especially in rural areas. Traditional grid reinforcements
are the preferred solution to increase the hosting capacity for PV at the LV level.
Table 26: Basic Information on the LV Level
Voltage
Basic Technical Guideline: Technical Guideline for the Connection and Parallel Operation of
Generators Connected to the LV Network [75]
Nominal line-to-line voltage
0.4 kV
Grid Layout
System topology
Vmax 3% [75]
Mostly radial, in urban areas
also meshed systems
Phase configuration
Five-wire/three-phase
Installation type
Mostly cables
Typical range of feeder length
100 m 500 m [85]
Load and
Generation
80
Figure 30: Example of a real German LV grid with high PV penetration. The system topology is radial.
81
reinforcement measures have to be borne by the DSO itself. The DSO is released from this
duty if the costs for the necessary reinforcement measures are economically unreasonable
9 (3) EEG. As of today, this case has not yet been specified in more detailed and is
considered to be highly case sensitive.
Technical Framework
The quality of the electrical supply in Germany is defined by the service reliability, the
voltage quality, and the service quality. As of today, maintaining the voltage quality is the
major issue when planning the interconnection of PV to the distribution grid.
The EN 50160 standard defines the required voltage quality for MV and LV customers
during normal grid operation in Germany. For the interconnection of PV systems especially,
the criteria for slow and fast voltage deviations have to be considered during the
interconnection planning process. Table 27 lists the requirements according to the EN
50160.
Table 27: Requirements for Voltage Deviation in German Distribution Grids According to EN 50160
Criterion
MV Level
LV Level
UN 10%
Reference
Value
10 min RMS
Evaluation
Period
1 Week
Slow Voltage
Deviations
Fast Voltage
Deviations
UN 10%
4% 6% UN
5% 10% UN
10 ms RMS
1 Day
Required Values
within Bandwidth
99% (MV)
95% (LV)
100%
82
Requirements
PN 100 kWp
30 kWp PN < 100 kWp
PN < 30 kWp
Required
Until
06/30/2012
31/12/2013
31/12/2013
Reference
6 and 66 EEG 2012
6 and 66 EEG 2012
6 and 66 EEG 2012
Although there has not yet been any uniform standard for the realization of the remote
control interface, most of the German DSOs use long-wave radio ripple control systems for
sending set values to the inverters.
Reactive Power Control:
Depending on the installed module capacity, PV systems have to provide a certain amount
of reactive power if demanded to do by the DSO. Table 29 lists the current requirements
according to the German MV and LV guidelines [74], [75].
Table 29: Requirements for Reactive Power Control for PV Systems
Voltage
Level
MV
System Size
Technical Requirements
Required Until
Reference
Any size
04/01/2011
LV
[74]
[77]
[75]
01/01/2012
01/01/2012
As of today, there is no uniform method for the provision of reactive power. For PV systems
connected to the MV level, the provision of reactive power can be either realized
autonomously (e.g., fixed preset power factor or voltage-dependent reactive power
provision Q(U)) or via remote control. For PV systems connected to the LV grid, only
autonomous (without communication link to DSO) reactive power provision methods are
used.
Chapter 0, best-practice approaches for determining the hosting capacity of certain grid
sections by taking into account the technical constraints of the grid operation as well as the
grid-supporting effects of active and reactive power control by inverters are introduced.
83
The reduction potential is determined by exemplary grid calculations; additional operational costs (e.g., grid losses,
maintenance costs) are not considered in the calculations.
84
Figure 31: Total net present value (NPV) of investigated local voltage control strategies referred to the
beginning of year t1. NPVInvest: Grid reinforcement, NPVOP1: Network losses + maintenance, NPVOP2:
reduced PV feed-in [84]
85
Table 30: Investigated Control Strategies of the HV/MV Voltage Transformer and PV Inverters in [90]
HV/MV transformer control
PV inverter control (MV level)
Fix voltage set point at the MV bus bar in the
Fix power factor of 1.0 (mode 1)
substation
Fix power factor of 0.95 (mode 2)
Power flow dependent voltage set point at MV
C
P
bus bar in the substation
Q(V) characteristic (mode 4)
* Mode 4 but with different slack-bus voltage. Shows sensitivity of Q(V) strategy on different
voltage magnitudes.
Figure 32 (left) compares the measured and simulated active and reactive power flow over
the HV/MV transformer for the summer day with alternating solar irradiation. In Figure 32
(right), a reactive power ramp rate sensitivity analysis at the HV/MV transformer is
performed for different autonomous voltage control strategies (provision of reactive
power) of the PV inverters for the same day. The reactive power control of PV inverters
leads to an increase of reactive power fluctuations in the grid, especially on days with
alternating solar irradiation. This effect can cause an increase of voltage variations at the
MV bus bar in the substation, which is the control variable of the HV/MV transformer
OLTC.
Figure 32: Measured and simulated power flow at a HV/MV substation (left) and reactive power
fluctuations at a HV/MV substation for different local control strategies of PV inverters (right) on a
summer day with alternating solar irradiation [90]
In [91], the parallel operation of a PV inverter and transformer control is evaluated in the
number of transformer tap changes and the maximum MV grid voltage. The impact of
reactive power control of the PV inverters on the number of transformer tap changes
differs significantly between the investigated control strategies.
The power flow dependent control of the HV/MV transformer improves the voltage
control in the grid but also leads to an increase of transformer tap changes
compared to the transformer control with a fixed voltage set point.
The voltage dependencies of the Q(V) characteristic have a damping effect on the
voltage step caused by the changes of transformer tap settings.
87
For the simulations, the central controller was assumed to rely on complete information
about the actual grid status. For real applications, this information would be limited to grid
nodes with measuring devices. Figure 35 shows the annuity of the required investments for
the central control plus operational costs (i.e., grid losses, Q penalties) in relation to the
amount of required measuring devices. The results show that the central control approach
could be cost competitive if the number of required measuring devices is low
Nevertheless, the central control approach has additional benefits for the DSO, which are
not explicitly considered in the presented analysis. In addition to reactive power
consumption for voltage support, coordinated active power control for an improved
congestion management and a detailed monitoring of the grid could be considered benefits.
Allocating the investment costs for the measurement devices on the different ancillary
services will further improve the economic benefit of the proposed control approach.
Figure 35: Annuity of investment plus annual operational costs for the scenario with a power factor of
0.95 (lagging) of the LV PV systems and for different PQ-bands at the connection point to the TSO [93]
89
90
Greece
Author:
Stathis Tselepis1
1Center
for Renewable Energy Resources and Saving (CRES), Photovoltaic Systems and
Distributed Generation Department, Athens, Greece
91
New PV grid-connected capacity increased rapidly over the past few years in Greece. The
new capacity for the year 2010 was 150 MWp, and for the year 2011 it was 400 MWp. In
the year 2012, due to the gradual maturity of many PV projects and the attractive feed-in
tariffs, and despite the economic situation, the annual PV installation capacity reached its
peak of 912 MWp. For 2013, despite the economic crisis, the momentum due to the
attractive feed-in tariff contracts that certain developers were still holding is expected to
boost the annual installed capacity to more than 1,000 MWp. The development of the
annual and cumulative PV capacity is presented in Figure 36. In the meantime, the Greek
government, under the pressure of the burgeoning renewable energy fund deficit, is
planning to further reduce the feed-in tariffs (possibly retroactively), while it has already
imposed a temporary (until July 2014) special tax on revenues on PV electricity generation
plants in Greece for PV plants connected to the grid in 2013 and before. Further, the
increasing penetration of PV systems should also push the authorities to study, plan, and
adopt technical and regulatory measures to allow higher penetrations of renewables in the
future.
2000
1600
1200
800
400
13
20
12
20
11
20
10
20
09
20
08
20
07
20
20
06
2400
YEAR
Yearly Installed Capacity
92
2800
1200
1100
1000
900
800
700
600
500
400
300
200
100
0
Value
2.564 GWp
2.428 GWp
9.894 GW
50.558 GWh
1487.8 GWh
2,94%
1,17%
39,66%
59,17%
2.564 GWp
2.428 GWp
As shown in Table 31, all the PV installed capacity is connected to the low and MV grid, and,
up to now, the PV-generated electricity is experienced by the system operator as a negative
load. This can be tolerated without any negative effects when the instant power
penetration level is relatively low and does not create any operation- and security-related
problems. Because there are already locations in the distribution grid where the local
penetration has increased significantly, measures are being taken at the moment to limit
new capacity in these grid segments by not allowing new PV installations.
Central and Southern Greece, where 65% of the country s electricity demand resides. The
Hellenic Electricity Transmission System consists of additional 400-kV and 150-kV lines, as
well as 150-kV submarine cables that interconnect Andros and the Western Greece islands
Corfu, Lefkada, Cephaloniia, and Zakynthos, and a 66-kV submarine cable connecting Corfu
to Igoumenitsa.
As of June 30, 2012, 11.303 km of transmission lines, 291 substations, and 619
transformers with a total installed capacity of 50.749 MVA comprised the capacity of the
Hellenic Electricity Transmission System [94].
Regarding the technical and regulatory framework, the code for the Hellenic Transmission
System management was completed in November 2012. The code does not consider
control or ancillary service provisions by PV systems connected at any voltage level to the
electricity grid. There is a general statement that the transmission system operator may
issue dispatch instructions for renewable energy and combined heat and power (CHP)
units; however, these deal exclusively with limiting production for reasons of safe
operation of the system, usually for wind plant, CHP, and hydroelectric plants.
The transmission and distribution of generated renewable electricity is guaranteed by
national law and the power purchase agreements (PPAs). In cases where renewable energy
plants are connected to congested areas, a predetermined upper limit of possible
curtailments is contracted; some remuneration of renewable energy plants is predicted for
such cases under L3851/2010.
Priority access (as defined by Directive
C to the grids is ensured as long as the
system security and security of supply are not jeopardized. Security criteria take into
account various factors and parameters such as:
One of the most important duties of the TSO (IPTO S.A.) is the daily energy planning of the
day-ahead scheduling, and one of tasks is the forecast of electricity injection from
renewable energy plants. For the day-ahead forecast of renewable energy generation in the
electric system, the main emphasis is given to wind plant energy production methodology
and correction. The other renewable energy technologies being considered are small
hydro, CHP, and biogas and PV plants. The next-day production forecasts of all the
technologies except wind make use of a model developed by the Aristotle University of
Thessaloniki that provides the energy that will be produced in the next day of each
renewable energy technology, except PV, for each hour of the day. Regarding PV systems'
production, the hourly energy production forecast of the previous day is taken into
consideration as well as the weather forecast for the next day and, more precisely, the
94
estimated radiation at selected locations, where PV systems have been installed and are
being monitored. From the monitored PV systems connected to the MV, daily production
data are collected. As to the PV systems connected to the LV grid, including PV systems on
buildings, the hourly daily output is deduced according to the latest figures of total energy
production provided by the DSO, DEDDIE (the Hellenic Electricity Distribution Network
Operator, or HEDNO) [95] for the time period of measurement, which is every four months.
Therefore, through this information, the change of installed capacity of PV systems is
calculated. Moreover, the transmission system operator receives daily irradiation data for
selected regions of the interconnected electrical system for the previous day and a forecast
for the next day. A large investment is being planned by the TSO regarding the installation
of a SCADA system in the distribution network.
National Distribution Grid Structure
The distribution network is radially operated. The nominal MV level is 20 KV, and the
nominal LV level is 380 V/220 V.
The key quantitative figures for 2011 of the Hellenic Electricity Distribution Network are
given below [95].
There are 228,900 Km of Network in total, including
107,500 km of MV network;
121,400 km of LV network;,
949 km of HV network, of which 205 km are in Attica and 744 km are in the noninterconnected islands;
224 HV/MV substations, of which 20 are closed-type substations, 199 are located
within the interconnected system, and 25 in the non-interconnected islands;
95
Technical requirements for the connection of independent generation to the grid, and
Additional provisions are included in the distribution network and transmission system
codes, as well as in the ministerial decree for the standardized grid-connection agreement
of DG stations.
The first directive comprises the technical evaluation framework and the limits set for DG
installations to be connected to the public LV or MV distribution network. Therefore, they
are DG-specific documents. Although the first one is applicable for any DG station, it still
comprises specific provisions for wind power stations. The second document concerns PV
installations specifically, with installed capacities up to 100 kW.
More specifically, the following technical considerations have to apply.
Voltage Regulation
The DG shall not actively regulate the voltage at the PCC. The voltage change induced by the
active power variations injected by the DG system must remain within 3% (10 minutes
average value).
Synchronization
Generators are connected to the network when voltage and frequency are close to nominal
values, or at least within the regulation limits suggested by protective means.
Synchronous generators shall use an automatic synchronization device with the following
characteristics:
o Maximum voltage difference: 10%;
o Maximum frequency difference: 0,5 Hz; and
o Maximum phase difference: 10.
In the case of many generators, the synchronization should occur sequentially with time
intervals longer than the transient times. Usually time intervals of a few minutes are
adequate.
Please note that the national rules conform the following international standards: HD 384 Electrical installations of
buildings; IEC; and CENELEC standards.
96
-20% to +15%
of nominal 230 VAC
50 0,5 Hz
Non-Interconnected
(Autonomous) Islands
-20% to +15%
of nominal 230 VAC
51 Hz to 47.5 Hz
In cases where the above window limits are surpassed, the inverter should be
automatically disconnected with the following time adjustments:
Further:
The DC injection in transformerless inverters should not exceed 0.5% of the nominal
or 20 mA for generators <16 A; and
Grounding
Grounding of the DG system shall be in accordance with HD 384-5-54/-55 and the relevant
national standards. The scheme of the DG grounding shall not cause over-voltages at the
equipment of DG or the equipment connected to the area EPS (electric power system). It
shall also not disturb the coordination of the ground fault protection of the area EPS.
For systems operating in grid-independent operation (intended island mode):
Protection by automatic disconnection of supply shall not rely upon the connection to
the earthed point of the public supply system.
When the DG system is operating in parallel with the electrical distribution network, there
shall be no direct connection between the generator winding (or pole of the primary
energy source in the case of a PV array or fuel cells) and the electrical distribution
network s earth terminal
97
Electrical Disturbances
The disturbances emitted by the DG system on the network must remain within the limits
described below:
The level of contribution of the installation to the distortion of the voltage must be limited
to values making it possible for the DNO to respect the acceptable limits regarding the
quality of the electricity delivered to the other users. The compatibility levels are given
below. They are valid for both LV and MV according to EN50160 and IEC 61000-3-6.
Table 33: Compatibility Levels for Maximum Harmonic Distortion in LV and MV Networks
The compatibility limiting levels for injection of harmonic currents to LV and MV grids are
given in Table 33. Lh is the ratio of harmonic current Ih (in A at MV) and short-circuit
power Sk (in MVA) of the installation at the PCC.
Flicker
The DER SYSTEM shall not induce objectionable flicker to other customers. National rules
conform to the IEC 61000-3-3, IEC 61000-3-7, IEC 61000-4-15, IEC 61400-21 standards.
Power factor
The output power factor must remain within a range agreed upon by the DNO. Additional
compensation measures may be taken.
98
Many types of power converters that consume considerable amounts of reactive power
may need compensation capacity. Self-commutated converters have small reactive power
needs and are usually capable of regulating their power factor between 0.95 leading to 0.95
lagging.
Safety
The generator shall operate safely over the entire designed and declared operating range.
For general requirements, EN61140 applies protection against electric shock are
common aspects for installation and equipment.
General requirements applicable to all operations of work activity on, with, or near
electrical installations are stated in EN 50110.
For the installation, the LV code must be taken into account. Warning labels must be placed
in live parts of the equipment according to Greek rules.
99
A remarkable renewable energy penetration case is in progress on the island of Crete [97].
Crete, with an annual peak load of 650 MW, is supplied by an isolated electrical system. The
annual renewable energy penetration for 2012 was approximately 20%, and the maximum
renewable energy capacity penetration 38.5% (280 MW of wind and 70 MW of PV).
During the early years of renewable energy operation, mainly with wind plants, some
problems were encountered due to the sensitivity of wind turbines to voltage dips, faults
on grid connection lines of wind plants to the HV substations (usually as a result of
contamination of insulators), and due to the voltage settings of the protective
disconnection devices of the wind plants. All of these problems were corrected by the DSO
and the wind plant owners, and the FRT capability of newer wind turbines especially and
impressively improved the performance of the system.
During the operation of the power system, PV systems provide their power output without
any restrictions, whereas wind plants contribute by also taking into account the allowable
instantaneous penetration levels, which are usually allowed to reach levels on the order of
40%. In cases when renewable energy penetration is higher than the allowable level, the
power of the various wind plants is equally limited so as not to exceed this level. The main
constraints for the instantaneous renewable energy penetration levels are determined by
the technical minimum of dispatched thermal units, the spinning reserve (N-1), plus
approximately 50% of the renewable power and additional network constraints
[100][101].
The Energy Control Centre of Crete continuously monitors the wind plants, and a set point
for maximum power output is given every 5 minutes in cases where this is required.
However, there are periods when the operators judge to operate the electrical system with
higher instantaneous renewable energy penetration levels (up to 60%), as shown in Figure
37.
The Energy Control Centre of Crete also selectively monitors online some PV plants at
various locations on the island, and by an up-scaling methodology the total PV production
is claimed to be calculated with a good accuracy, which helps in the daily scheduling of the
conventional plants. The distributed PV plants have also been shown to support the
maintenance of the voltage level in the acceptable range throughout the grid during the
daily hours. The experience in island grids has generally shown no increase in the amount
of required reserve capacity but increases in the use of operating reserve (regulating
power 10 min. 15 min).
100
Figure 37 G
which allows net-metering for PV systems of all sizes and small wind turbines for autoproducers in Greece. The PV systems that make use of net-metering are exempt from any
suspension of the licensing process. The energy produced by the PV plant is fed into the
distribution network and is subtracted from the energy consumption of the auto-producer
in each count cycle (every four months). Any excess energy provided to the grid resulting
from the previous offsetting procedure receives no compensation. Especially for public
bodies that install PV systems in the framework of European programs, it is already
provided that up to 20% of the total energy produced per year may be compensated by the
respective feed-in-tariffs.
102
Italy
Authors:
Adriano Iaria,1 Antonio Gatti,1 Diego Cirio1
1Ricerca
sul Sistema Energetico RSE SpA, Energy Systems Development, Milano, Italy
103
At the end of year 2012, the installed PV capacity in the Italian power system was 16.6 GW;
18.3 TWh of energy were supplied by PV [103] on a total energy demand of 325 TWh. The
PV penetration is very high, taking into account that in 2012 the Italian peak load was 54
GW. Table 34 summarizes the recent statistics about PV generation and power
consumption in Italy.
Table 34: Statistics on PV Generation and Power Consumption in Italy
Statistic
Installed PV capacity (nationwide)
Peak load (nationwide)
Total generation capacity
Total energy generated by PV in 2012
Share of PV on total energy consumption in 2012
Share of installed PV capacity connected to HV level
Share of installed PV capacity at HV/MV substations
Share of installed PV capacity connected to MV level
Share of installed PV capacity at MV/LV substations
Share of installed PV capacity connected to LV level
Average size of PV system
Value
17.1 GW
54.1 GW
128.1 GW
18 862 GWh
6.1%
5.8%
n.a.
63.3%
n.a.
31.0%
34.3 kW
Year
Status: 08/2013 [104]
Status: 31/12/2012 [103]
Status: 31/12/2012 [105]
Status: 31/12/2012 [105]
Status: 31/12/2012 [105]
Status: 31/12/2012 [106]
Status: 31/12/2012 [106]
Status: 31/12/2012 [106]
Status: 31/12/2012 [106]
A rate of 96% of the installed plants (458,265) is connected to the LV level; only 20,000 PV
plants are instead connected on the MV level, but they represent the 63.3% of the total
installed power capacity [106]. Only 229 PV plants are connected to the HV level. Figure 38
depicts the allocation of the installed PV plants in the 20 Italian regional areas (year 2012
[106]).
104
The distribution MV and LV networks are radially operated and consist of 1,232,000 km of
lines (year 2012 [107]): 385,284 km at the MV level; 846,507 km at the LV level.
MV Level: The MV distribution grid mainly covers the standard range from 15 V to 20 kV,
with few exceptions included in the larger range from 5 V to 23 kV. The rated capacity of
HV/MV transformers in primary substations is usually between 16 MVA and 63 MVA, but
some exceptions are in the larger range from 10 MVA to 100 MVA. The MV level is realized
as a three-phase star system operated in a radial scheme either in urban or in rural areas.
The star point is ungrounded or grounded by means of Petersen Coil aimed to limit the
arcing currents during ground faults. The technical Italian standard CEI10 0-16 [108]
establishes the limits for active/reactive power exchange and control.
LV Level: The three-phase with neutral conductor LV distribution level has a nominal
voltage of 400 V/230 V. The rated capacity of Delta-Star MV/LV transformers, in
secondary substations for energy supply to households is typically from
kVA to
630 kVA, whereas higher rates (typically 1,000 kVA, 2,500 kVA) are possible for MV
industrial users or energy producers The general practice of )talian distributors doesn t
include the remote active power control of DG connected to the LV level. However, the
recent Italian standard CEI 0-21 [109] establishes that, in the case of the availability of a
proper communication system, remote control should be implemented for generating
plants that have a rated power higher than 6 kW.
10
11
The LV level will be considered in cases with nominal generating capacity up to 100
kW;
106
On the basis of technical reasons, the grid operator may supply the connection
service at the LV and MV levels in cases with generating capacities respectively
higher than 100 kW and 6,000 kW.
Connection costs have to be borne by the plant operator with the exception of extra costs in
cases where the DSO choses a connection solution that is more expensive than the
minimum technical one; in the latter case, the DSO has to pay for the extra costs. The TICA
procedure also provides guidelines concerning the cost estimation of connection.
Technical Framework
According to the above AEEG resolutions; the CEI standards are based on EN 50160 and
represent the main national technical references for connections on the distribution
network. In particular, CEI standard 0-21 [109] (last updated in 2012) sets up the technical
rules for connection on the LV level; whereas CEI standard 0-16 [108] (last updated in
2013) states the rules for connection to the MV level. The requirements for DG connection
as set out by CEI 0-16 and CEI 0-21 have also been endorsed in a recent annex (A70 [113])
of the grid code of the Italian TSO.
107
PN
kW
PN
kW
PN
kW
PN
kW
PN
kW
PN
kW
PN
kW
PN
kW
PN
PN
kW
kW
Date
of
V Level Reference
Connection
Requirements
Remote control, in case remote signals are
available.
Operation
range
ensured:
V V
Vn
H
HV
H
(in case of transients due to MV level)
Remote control, in case remote signals are
available.
Operation
range
ensured:
V V
V
H
HV
H
(in case of transients due to MV level). Active
power modulation with frequency droop 2.4%
in the overH
LV FRT (LVFRT) capability if rated power no
less than 6 kVA.
Operat
H
V 110%Vn;
Operation range
V V
V
H
on the basis of remote protection control
availability.
Operation range
V V
V
H
on the basis of remote protection control
availability. Active power modulation (freq.
droop 2.4%) in the over-frequency range
H
LVFRT
power no less than 6 kVA.
(retrofitting) Remote control, in case remote
signals are available. Operation ranges
V V
V
H
(in case of transients due to HV level) or
H
MV level)
(retrofitting) Operation range ensured:
V V 110%Vn
H (By 2014 June
th
30 )
(retrofitting) Operation range ensured:
V V 110%Vn
H (By 2014 June
30th)
(retrofitting) Operation range ensured:
V V 110%Vn
H (By 2015 April
30th)
108
From
01/04/2012
to 30/06/2012
MV
From
01/07/2012
MV
From
01/04/2012
to 30/06/2012
LV
From
01/07/2012
to 31/12/2012
LV
After
31/12/2012
LV
Before
01/04/2012
MV
Before
01/04/2012
MV
5bis
243/2013
Before
01/04/2012
LV
5bis
AEEG
243/2013 [114]
Before
01/04/2012
LV
5bis
AEEG
243/2013 [114]
AEEG
Table 36 lists the current requirements according to AEEG resolutions number 33/08 and
187/11 and the CEI standards 0-16 and 0-21.
Table 36: Requirements for Reactive Power Control of PV
V Level
System Size
MV
PN <400 kW
MV
PN 400 kW
LV
PN
LV
3<PN
LV
PN > 6 kW
kW
kW
Technical Requirements
Minimum power factor of 0.9
(leading/lagging) and
rectangular generator/inverter
P-Q capability.
Minimum power factor of 0.9
(leading/lagging) and
semi-circular generator/inverter
P-Q capability.
Minimum power factor of 0.98
(leading/lagging)
Minimum power factor of 0.95
(leading/lagging)
Minimum power factor of 0.9
(leading/lagging)
Reference
AEEG 33/2008
8.8.5.3 CEI 0-16
[108]
AEEG 33/2008
8.8.5.3 CEI 0-16
[108]
AEEG 187/2011
8.4.4.2 CEI 0-21
[109]
AEEG 187/2011
8.4.4.2 CEI 0-21
[109]
AEEG 187/2011
8.4.4.2 CEI 0-21
[109]
109
Figure 39: Hosting capacity in each MV node. Cumulative histogram of MV node percentages with the
concerning installable DG powers [117]
In general, it was confirmed that less-stringent limits may be achieved by means of new
measures on the protection system and the grid automation in addition to a more efficient
management of the DG interface protections as stated by new standards mentioned above
(CEI 0-16 [108]; Annex A70 of Italian Grid Code [113]).
The second study was performed on the LV distribution network starting from a network
data collection (years 2007 2008). It concerns a reduced set of the distribution LV network
consisting of 500 MV/LV substations. A further reduced subset of 16 MV/LV substations
was then analyzed to assess the DG (mostly PV) hosting capacity on the LV level. The subset
represents only the
of the national LV network; nevertheless, it was properly built
with the aim to represent the most common configurations and situations. Considered
constraints included slow voltage deviations within 10%UN (EN 50160), line capacity
based on the conductor thermal limit, and fast voltage deviations from 5% to 10% UN. The
cumulative histogram in Figure 40 ([118]) shows that most of the nodes (83% of total
analyzed) allow an installed DG of 30 kW; whereas 25% of the nodes allow an installed DG
of 100 kW. Fast voltage deviations under 5% UN are the most stringent constraints because,
as shown in Figure 40, they affect the main share of nodes involving constraint violations.
110
Figure 40: Hosting capacity in each LV node. Cumulative histogram of LV node percentages with the
concerning installable DG powers [118]. Maximum fast-voltage deviations 5%
According to EN 50160, for fast voltage deviations, a higher limit of 10% UN may be
assumed. Figure 41 shows that, with this larger limit, fast voltage deviations become a nonstrict constraint: higher percentages of nodes 95% and 45%, respectively allow 30 kW
and 100 kW of installed DG.
Figure 41: Hosting capacity in each LV node. Cumulative histogram of LV node percentages with the
concerning installable DG powers [118]. Maximum fast voltage deviations 10%
Instead, line transfer capacity becomes a relevant constraint that can be faced with
expensive network developments. The maximum short-circuit current in each LV line was
found as a not much binding constraint; in any case, such a problem may be overcome by
means of the increase of MV/LV substation numbers with the adoption of smaller size
transformers (rated power lower than 630 kVA).
The study points out how a smart grid conception may facilitate the grid operation e.g., in
cases of fast voltage deviations, unintended DG tripping can be prevented by selective
protection logics; slow voltage deviations may be faced with local voltage control by DG. In
general, a proper communication system with a remote controlled interface of DG may help
avoid local and system problems.
111
To reach a better quality of supply and a higher PV hosting capacity, local voltage control
aimed to face slow voltage deviations is already foreseen by the recent standards CEI 0-16
(MV-level [108]) and CEI 0-21 (LV level [109]). These standards state that the distributors
may ask for the adoption of the following voltage control logics:
; power factor as function of injected active power P to limit overvoltages due to active power injections;
These control logics have to be applied accordingly within the control capabilities shown in
Table 36. Implementation of remote signals by DSO to achieve a coordinated DG voltage
control is currently under discussion.
Within the National Renewable Energy Action Plan (adopted June 2010), this RES-E energy
quota was split into the following major energy sectors: heat, transport, and electricity. The
contribution of renewable electricity generation was fixed at 26.4% of gross electricity
consumption for 2020 (corresponding to about 100 TWh/year of renewable production).
According to the National Renewable Energy Action Plan [119] (see also Art. 4.1 of
2009/28/EC [120]), the 2020 target of solar installed capacity was fixed at about 8.6 GW
(including 0.6 GW of solar-thermodynamic technology). To reach this objective, a feed-in
tariff for PV systems was implemented (ceased with the last PV Support Mechanism, named
V Conto Energia, May
As of December 2012, the cumulative PV installed capacity exceeds 16.4 GW so that,
including other renewable energy production (hydro, wind, biomass, and geothermal), the
total renewable energy production covers about 27% of gross electricity consumption.
Further solar development is included in the last Italy National Energy Strategy (March
2013 [122] in grid parity condition about GW year up to
To reach the new
national renewable energy goal for 2020 of 35% to 38% of gross electricity consumption
(about 120 TWh to 130 TWh/year of renewable energy production), the allocated
investment resource is about 11.5 bln year to12.5 bln year for
years To implement
this energy strategy, new energy support mechanisms are currently under discussion.
112
Japan
Authors:
Yuzuru Ueda,1 Kazuhiko Ogimoto,2 and Koji Washihara3
1Tokyo
3New
113
Value
4.9 GW
156 GW
287 GW
3,890 GWh
0.4%
20%
Year
Status: 03/2012 [123]
Status: FY2012 [124]
Status: 03/2013 [124]
Status: FY2012 [125]
Status: FY2011 [125]
Status: FY2011 [123]
80%
<5 kWp
114
116
Table 38: Outline of the Technical Requirements Guideline of Grid Interconnection to Secure
Electricity Quality in Japan (the criteria applicable only to PV) 1; *1: Specific case denotes that the
capacity of the inverter is less than 2 kVA in 1p2w, less than 6 kVA in 1p3w, and less than 15 kVA in
3p3w, or that the power factor at the point of common coupling is usually close to 1, such as in
residential systems. *2: ex. single-phase two-wire)
Network
Definition of
the network
LVJ
LV distribution
network includes
single-phase twowire (1p2w) 100 V,
1p3w 100 V/200 V,
3p3w 200 V and
3p4w 100 V/200 V
HVJ
HV network to
supply electricity
to larger
consumers and
distribute the
power from
distribution
substation to pole
transformers.
3p3w 6.6k V is
typically used.
Spot network
More than two
lines are supplying
in parallel to the
consumer. Typical
voltage is 22 kV
and 33 kV.
Generator
capacity
Power factor
(In case
reverse
power flow is
permitted)
< 50 kW
< 2,000 kW
< 10,000 kW
Automatic
load
limitation
Voltage
fluctuation in
normal
operation
EHVJ
HV transmission
network to supply
electricity to
larger consumers
and transmit the
power to
substations.
Below 35 kV can
be classified as
distribution
network in some
cases.
2,000 kW
117
Table 39: Outline of the Technical Requirements Guideline of Grid Interconnection to Secure
Electricity Quality in Japan. (These criteria are applicable only to PV.)
Network
Instantaneous
voltage
fluctuation
(selfcommutated
inverter)
Instantaneous
voltage
fluctuation
(externally
commutated
inverter)
Prevention of
unnecessary
disconnection
LVJ
HVJ
Use automatic synchronizing function.
Islanding
Not allowed.
Spot network
EHVJ
As the same as
others if more
than 2% of the
voltage deviation
is expected.
Not disconnect, if
the duration of
the voltage drop
is shorter than the
specification.
Not disconnect, or
recover within the
recloser s reaction
time without
affecting the
reverse power
relay and
underpower relay,
if the fault occurs
in the nonconnected line.
Not disconnect, or
recover within the
recloser s reaction
time without
affecting the
reverse power
relay and
underpower relay,
if the fault occurs
in the nonconnected line.
Disconnect the
fault line but stay
connected with
other lines.
Generator should
not be
disconnected.
Islanding
operation is
allowed with OFR
and UFR or
remote breaker.
By complying with the ordinance of the METI, PV systems shall be interconnected to the
grid under the technical framework of the Grid Interconnection Code (JEAC9701). The
outline of JEAC 9701 is described in Table 40. The requirements for grid interconnection
are classified into the requirements for LVJ distribution lines, HVJ distribution lines, and
EHVJ transmission lines, respectively. Interconnection to a 200-V/100-V distribution line is
designed for rooftop application (residential small PV system), whose generating capacity
is below 50 kW (typically several kW). Interconnection to a 6.6-kV distribution line is
designed for industrial or business use PV systems whose capacity is between 50 kW and 2
MW. Interconnection to a 66-kV or higher voltage transmission line is designed for large PV
systems (mega-solar) that have a capacity larger than 2 MW.
118
As for inverter certification scheme for grid interconnection, Japan Electrical Safety &
Environment Laboratories (JET) provides a certification program for Grid-connected
Protective Equipment for Small Distributed Generation Systems and certifies inverters
with capacity less than 20 kW for interconnection to the LV network. This certification
program aims to enable smooth grid interconnection by both utilities and distributed
power producers. No certification scheme is established yet for the inverters of industrial
or business use (larger than 20 kW). In this case, a PV system installer is required to have a
technical consultation with the competent utility and ensure conformity to JEAC9701.
Table 40: Outline of Grid Interconnection Code (JEAC9701). *1. Standard active islanding detection
scheme: A frequency feedback method with step reactive power injection (JEM 1498 (2012 JEMA)).
See 2.8.3. *2. Voltage Q (reactive power) control
HVJ distribution line
interconnection
6.6 kV
50 kW to 2 MW
More than 0.85 lag
Autonomous control at
DER
Reactive power and
active power control
-
Short-circuit current
Protection
-Internal fault
OCR, OCGR
OCR, OCGR
-DER outage
-Grid fault
OVR, UVR
DSR, OVGR
OVR. UVR
DSR,OVGR
Anti-islanding
OFR,UFR
Anti-islanding
Mechanical
disconnection
Generating capacity
Power factor
Voltage fluctuation
Disconnection method
119
Restriction act is
needed.
If the voltage is over the regulated value and the inverter is equipped with the reactive
power control function, the inverter should perform reactive power control within the
range of 1 to 0.85 p.f. through lead angle operation.
If the voltage cannot be within the regulated value by reactive power control or the
reactive power control is not equipped, an inverter should perform active power
control until the voltage becomes the regulated value.
The autonomous control procedure required for HVJ interconnection is basically the same
as the LVJ interconnection, except for the following items:
The regulated voltage value is determined through discussion between the DG owner
and utility.
Reactive power control is perfromed by both power factor control and Static Capacitor
(SC ) control.
Anti-islanding:
A single-phase inverter in LV use must have both passive and active islanding detection
functions in addition to OVR, UVR, OFR, and UFR. For the active method, the new standard
anti-islanding protection method is defined in JEM1498 by Japan Electrical Manufacturers
Association (JEMA). The principle of the method is the frequency feedback method with
step reactive power injection, which is called JSIDF (Japanese standard islanding detection
function).
120
The JSIDF for high-penetration PV systems was developed with the following concepts.
1) High-speed islanding detection
2) Without mutual interference of active signals from multiple inverters
3) Without misdetection for frequency change or sudden voltage drop due to
electric accident in bulk power systems.
There are two types of active signals in JSIDF. The algorithms of active methods are as
follows:
1) Frequency feedback function
Frequency usually changes naturally after utility stops power distribution because of the
imbalance in demand and power generation. Frequency feedback function injects reactive
power to increase the frequency change to detect islanding. Reactive power is injected
synchronously from each inverter because the same frequency value can be detected at
each inverter. The correlation between frequency change and injected reactive power is
shown in Figure 43.
Reactive power injection
Second gain
Frequency
change
First gain
The first gain is set smaller than the second gain to reduce reactive power injection during
a normal power grid condition. The second gain was set to larger value to collapse the
reactive power balance. The threshold frequency between first and second gain is +/- 0.01
Hz. The maximum absolute reactive power is 0.25 p.u.
2) Step injection function
Step injection function is applied to assist the frequency feedback function when frequency
change does not occur after the power distribution stops under the power balanced
condition. We focused on the harmonics distortion of voltage because it varies due to the
characteristics of transformer.
121
Step injection function injects reactive power in case the total harmonics distortion (THD)
voltage increases over 2.5 V and continues three cycles and frequency does not change
during these three cycles.
The reactive power is injected only two cycles. If frequency changes after the two-cycle
injection, frequency feedback function will be operated. From the 2nd to 7th voltage
harmonics is used to compute THD.
Fault-Ride-Through:
FRT is required for DES (distributed energy source) so as not to be disconnected from the
grid at the time of temporary voltage drop due to the fault at transmission line or load
switch over operation. As is the high penetration of PV systems realized, FRT becomes an
essential function to ensure the stability of power system. If most of PV systems are
disconnected from the grid under such conditions, large-scale power supply failure will be
caused and the power system will get unstable.
As described in 2.8.2, grid-interconnection requirements are classified into the voltage
classes of distribution lines or transmission lines to which the DES is interconnected. But
FRT, being di-electric duration capability of DES itself, needs to be specified by the kind of
DESs. The necessary FRT function required for PV systems had been discussed and
authorized at first among every type of DES.
Figure 44 shows the FRT requirements for PV systems. The requirements are classified into
the requirements for a single phase inverter and a three-phase inverter. When you test the
FRT function of a three-phase inverter that is used for PV systems of large capacity, cost or
space factor should be considered to lead to practical testing procedures.
FRT requirements are to be applied step by step from the tentative specification stage to
the final specification stage, considering technological development. FRT requirement for
PV systems is being translated and transferred to other types of DERs, that is to say, battery
systems, fuel cell systems, and gas engine cogeneration systems.
122
LV duration
capability
3 phase inverter
[Tentative] until Mar.2017
1sec
0.3sec
100%
Ride through
30%
30%
Ride through
or Gate block
Ride through
Voltage
Voltage
100%
Ride through
or Gate block
1sec
100%
Ride through
20%
0.3sec
Voltage
Voltage
100%
Ride through
or Gate block
Response time to
[Tentative] until Mar.2017
return to 80% of rated
within 0.5 sec
voltage
[Final] after Apr.2017
within 0.1 sec
Ride through
20%
Ride through
or Gate block
within 1 sec
Figure 45: Schematic of the supply demand balance throughout 24 hours in a system with a large
number of PV systems
The NG-TDN project deals with the above issues by setting four groups according to
technical fields to provide solutions to massive penetration of PV systems:
Group 1: Methodology for optimum allocation and control of voltage regulating devices
124
LPC-P
LBC-P
LPC
LPC-Q
LBC-Q
0
-50
-100
-150
-200
-250
-300
0
1000
2000
3000
4000
Time [sec]
6900
Voltage [V]
6800
Upper Limit
6700
6600
6500
6400
0
1000
2000
3000
4000
Time [sec]
Demonstration tests of the developed centralized voltage control system using voltage
control devices such as SVC, LPC, etc., in ADAPS were carried out at the Akagi testing center
of CRIEPI [129]-[131]. The experimental distribution system shown in Figure 48 (a)
consists of a substation, 6.6 kV distribution lines, several line voltage sensors,
communication network, three units of 100 kW simulated DG, three units of 200 kVA load,
and the developed voltage control devices. Each DG is the inverter type generator
simulating PV system. Two distribution lines of feeder 1 and feeder 2 are connected by the
LPC.
Figure 48 (b) and (c) show typical test results using a 1,000-kVA LPC. Figure 48 (b) shows
the LPC active power (P) and reactive power (Q) changes according to DG generation
power and load consumption. The LPC absorbs surplus the active power of DG in feeder 2
and injects the power to feeder 1. As a result, the distribution line voltage of each location
in feeder 2 is stably controlled under the upper limit shown in Figure 48 (c). Required
device capacity and distribution line power loss can be generally reduced via a LPC
compared to SVC [131].
127
Using Japanese average distribution line models, proper line voltage control measures
according to the penetration rate of DG to distribution line capacity were evaluated by the
means of demonstration tests and numerical simulations, taking account of the ability to
control proper voltage range and system cost [129][131]. Results from the case in which
DG is penetrating the LV distribution feeder are shown in Figure 49. In residential areas, a
designed centralized control method using SVC or LPC using information of the whole
power distribution system may be needed when the penetration rate reaches 65% or more.
As the results show, the possible penetration rate in a feeder can reach 100% by the
designed voltage control method combined with reactive power control of a PV system.
Figure 49: Evaluation results of proper voltage regulation methods according to DG penetration rate
and discharge plans for storage batteries including electric vehicles (EVs) for utilizing
surplus PV power.
The demonstration test to evaluate the developed method was carried out using a 4-kWp
PV system, 1.0 kW (3 kWh) HPWH system, and other residential electrical appliances [133].
Figure 51 shows an example of PV and load profiles on a clear day when the reverse power
limit is set at 2 kW. About 50% of the expected PV generation loss can be utilized by
daytime operation of HPWH according to the plan calculated by the developed method. As
the results of year-round operation, average daily PV generation loss is significantly
reduced from 1.34 kWh to 0.65 kWh. The reduction rate of average yearly electricity
charges per customer reaches a maximum of 20%.
For future work including the ADAPS techniques, new coordinated operation and control
methods of distributed smart communities and the whole power system will be
investigated.
129
PV generation
130
Spain
Author:
Manoel Rekinger1
1European
131
Value
4682 MW (AC)
39640 MW
104159 MW
8,1 TWh
2.99%
17%
48%
36%
107 kW
Year
November 2013 (UNEF)
December 2013 (ENTSO-E)
December 2013 (ENTSO-E)
January 2013 (EPIA)
January 2013 (EPIA)
January 2013 (EPIA)
January 2013 (EPIA)
January 2013 (EPIA)
January 2013 (EPIA)
The grid codes or operational procedures (POs) required by the Spanish TSO, Red
Electrica de Espana; and
The Control Centre of Renewable Energy, CECRE, which has been in operation since
2006. CECRE is considered a worldwide pioneering initiative to monitor and control
renewable power plants, specifically wind farms.
Taking into account the penetration of renewable energy in the Spanish power system,
some POs apply to wind power and large PV systems, but their focus is on the prevention
and resolution of system-related disturbances. This is what makes the Spanish situation
unique in the European landscape compared to countries like Germany or Italy. In these
countries the focus of grid code evolution has been to address mainly distribution-level
challenges. However, it is worth mentioning the main characteristics of grid connection
requirements for PV systems in Spain, as they could interact12 with distribution systems
operation or be extended to small PV systems:
P.O.12.3/R/D 1565: FRT requirement (zero-power-mode) for all generators larger than
2 MW
P.O. 12.3: Reactive current injection at FRT for all new generators larger than 2 MW
12
Positively; the capabilities could be used for distribution system operation or negatively; their use could trigger
new challenges at the distribution level.
132
Submission of real-time data (V,P,Q) to the CECRE for all new generators larger than 1
MW
Voltage support though a Q(V) static function and frequency support though a P(f)
function for all new generators larger than 10 MW.
Figure 52: Overview of the networks considered in the case studied: a MV grid (left part of the figure)
and a HV grid (right part) (Source: ReserviceS project 2014 [135])
133
The main characteristics of the two considered networks are given in Table 42.
Table 42: Characteristics of the Two Networks Considered for the Analysis. (Source: ReserviceS project
2014 [135])
MV network
15 kV
Generation capacity
Peak Load
Lines length
Installed PV capacity
Installed Wind capacity
2.311 MW
7MW
23 km
0.331 MW
0 MW
HV network
45-66-132-220-400
kV
1,300 MW
3,000 MW
6,000 km
3.5 MW
384 MW
Different penetration scenarios for wind and PV have been taken into account during the
analysis of the two voltage control strategies. These scenarios are presented for the MW
network and for the HV network.
Table 43: Wind and PV Scenarios Envisioned for the Analysis of the Voltage Control in the MV
Network (Source: ReserviceS project 2014 [135])
MV Scenarios
Current situation (DG1)
Wind
PV
0.311 MW LV
PV increase (DG2)
5 x 0.2 MW MV
2 x 0.2 MW MV
REserviceS scenario (DG4)
Maximum DG (DG5)
Other
Demand
Generation
2 MW CHP
Peak
0.311 MW LV 2 MW CHP
5 x 0.2 MW MV
0.311 MW LV 2 MW CHP
Peak
0.311 MW LV 2 MW CHP
3 x 0.2 MW MV
Peak
5 W x 3.5 MW MV
0.311 MW LV
2 MW CHP
Peak
Peak
Table 44: Wind and PV Scenarios Envisioned for the Analysis of the Voltage Control in the HV Network
(Source: ReserviceS project 2014 [135])
HV Scenario
Current
situation
(DG1)
Wind
PV
357 MW HV
3.5 MW HV
134
Other
Generation
1492 MW HV
Demand
Off-peak
DG2
DG3
DG4
1000 MW HV
8 MW HV
1280 MW HV
Off-peak
357 MW HV
204 MW MV
710 MW HV
510 MW MV
3.5 MW HV
1290 MW HV
Off-peak
10.4 MW HV
630 MW HV
Off-peak
Moreover, other assumptions have been taken regarding the average R/X ratio of the MV
network:
R/X1: is the value of the R/X ration of the analyzed MV network (R/X avg = 1.5)
Finally, different distributed generator behavior models have been developed depending
on the type of technology. Wind turbines have been assumed to be equipped with a
STATCOM enabling an unlimited Q provision at the point of common coupling. For MV
connected PV systems, a limited Q provision with a cos of
has been assumed Finally,
no Q provision is assumed for LV connected PV systems and the other generators.
The DSOs involved in this analysis took also other assumptions beside the penetration
levels, the R/X ratios, and the reactive power provision models for the DG. A high-level
description of the methodology used to conduct the analysis and the assumptions is
presented in Figure 53.
135
Figure 53: High-level description of the methodology and assumptions used for the analysis (Source:
ReserviceS project 2014 [135])
Methodology
Two optimal power flow (OPF) analysis have been conducted using two different
constraints:
V pu
on the two R/X rations assumptions and the five (for MV) and four (for HV) penetration
scenarios considered. As shown in Figure 54, the two OPF were performed to analyze four
different impacts of DG penetration on the networks.
136
Figure 54: Scope of the analysis (Source: ReserviceS project 2014 [135])
The effectiveness of the reactive power - voltage control is calculated as the maximum
voltage variation obtained by means of the reactive power injection or withdrawal needed
for both OPF solutions. Two elements will be analyzed:
Qgen: the reactive power provided (injected or withdrawn) by the generators (unit of
measurement: Mvar)
The impact of Q-V control on active power losses is calculated as the maximum active
losses reduction (referred to BAU DG1 scenarios) obtained by means of the voltage control
defined with the collaboration of DG. Two elements will be analyzed:
Qgen: the reactive power provided (injected or withdrawn) by the generators (unit of
measurement: Mvar)
Qgen: the reactive power provided (injected or withdrawn) by the generators (unit of
measurement: Mvar)
The voltage violation will be represented with the minimum and maximum voltage value
in any node (Vmin,Vmax):
137
Figure 55: Results of the voltage violations analysis for the R/X2 MV network (Source: ReserviceS
project 2014 [135])
The results show that even in the case of a weak network (R/X2), a smart voltage control
approach based on Q set points sent to the MV connected generators can maintain the
maintain the voltage within the defined boundaries if enough controllable capacity is
installed. For instance, in the DG2 scenario, the five 200 kW PV systems cannot compensate
the voltage rise at the MV bus bar.
Impact on P losses:
DG active power injection may cause a negative or positive effect on the active power
losses depending on its penetration level. Q-V control provided by the generators is
effective as long as losses reduction objectives are taken into account. On the other hand, if
Q-V control is used to avoid voltage limits violations, losses may even increase.
138
Figure 56: Impact of voltage control on active power losses (Source: ReserviceS project 2014 [135])
Figure 57: Effectiveness of Q-V control depending on the R/X ration and the scenario (Source:
ReserviceS project 2014 [135])
139
Figure 60: Reactive power generation for the different considered scenarios (Source: ReserviceS
project 2014 [135])
Cost-benefit analysis of Q-V control in MV networks
Preliminary remark
All the results presented in this section are based on assumptions related to different
simulations and not to any concrete project or existing network portions. The aim of this
simulation is to show how the implementation of new functionalities could bring benefits
at the system level, using smart grid technologies within a network configured as in the
case studies presented before. As demonstrated in this analysis, the result is case
dependent, varying significantly when applied to different scenarios.
Key findings
On one hand, costs occur due to new investments. The main assets considered in the cost
benefit analysis are the following:
Devices to control DG active (Active Power Regulator or APR) and reactive power
SCADA upgrade
As shown in the case study, the proposed voltage control strategy can bring new benefits.
The following benefits have been estimated and considered:
142
Moreover, Q-V control solution would enable an increase in DER hosting capacity so that
indirect and external benefits, not monetized in this cost-benefit analysis, could also be
taken into account; among these indirect benefits, one could consider the:
The CBA results are presented in Table 45 [135] and are based on a 20-year net present
value analysis with 2013 price references.
Two different case studies have also been considered with the same MV network, the same
demand/generation forecasting, but different characteristics of the lines:
Weak line scenario where voltage problems cause outages and curtailment to be
solved
Present line scenario, where the absence of voltage problems drives loss minimization
strategy.
Table 45: Results of the CBA Methodology Applied to Two Different Network Conditions. (Source:
ReserviceS project 2014 [135])
As represented in Table 45, even though some external benefits have not been monetized,
in both cases the investments and costs are compensated by the monetized benefits, which
lead to a positive net present value . This means that the smart solution for DG integration
13
Although curtailment is not allowed in Spanish regulation frame, it has been considered for this use case to extract
information about benefits regarding curtailment.
143
is, for these two specific cases, cheaper than BAU. But different quantitative impacts are
identified.
Not also the final result, but also the distribution of the investments and costs differs
between the two cases, resulting on some cases on negative impact on the DSO or the DG
side. Due to the differences between both cases, the distribution of savings among the
different expected benefits also varies.
144
Switzerland
Authors:
Davy Marcel1 and Christof Bucher2
1PLANAIR
2Basler
145
Positioned in the center of Europe, Switzerland plays a leading role in the European
electricity market. About 11% of the all the produced electricity in Europe flows through
Switzerland, making it an important transmission country. Figure 1 illustrates the energy
flows with neighboring countries for the year 2012.
The Swiss power plant infrastructure as a whole is characterized by a high level of capacity
for peak load. This capacity covers the higher day-time consumption, in other words the
peaks in demand in the morning, at midday, and in the evening, and evens out short-term
imbalances between supply and demand. Peak load energy places higher technical
demands on production, as either storage or flexible power plant deployment is required.
Figure 61: Overview of Swiss import/export and transmission in 2012 (Source: Swissgrid)
146
Figure 62: Overview of Swiss transmission and distribution grid (Source: Swissgrid)
The grid level 1 is an EHV transmission system that is fully owned and managed by the
unique Federal TSO Swiss grid since 2013. The system operates with 380 kV or 220 kV to
reduce the amount of energy that is lost due to transmission. It is used for transmitting,
importing, and exporting electricity and distributing electricity to lower grid levels. The
total span of the transmission system is 6,700 km.
The national Distribution grid system in Switzerland is very fragmented in terms of
responsibilities and number of actors involved (for historical and structural reasons).
There are remaining around 800 distribution grid managers (Source: Swiss grid internet
site), who are acting on a territory of 47,285 km2, each of them responsible for very various
areas in terms of size, going from villages or towns up to inter-cantonal and regional scales.
The Supra-regional grid level (between level 2 to 4) uses voltage levels from 150 kV to 36
kV and is used to supply energy to cantons, regional, and urban distribution grids as well as
large industrial areas. The distribution grid represents the majority of the 250,000 km of
the entire grid length and is 80% underground (Source: AES/VSE internet site / grid facts).
The supra-regional grid is owned and managed by several actors, often having also roles as
producers and/or DSOs. Examples of major DSOs that may be named are EWZ, EKZ, BKWFMB, Axpo, SIG, Groupe e, Romande Energie, REPower, CKW, FMV, EWO, AET, IWB, etc. All
147
these actors own and/or manage various grid levels. Some actors are producers (e.g.,
Alpiq), some of them are not owners of their grid. Many system voltage levels can be found
(e.g., 50 kV, 60 kV, 65 kV, 125 kV, 132 kV, 150 kV), and it must be expected that protection
(mechanisms, relays performances) as well as network grounding (direct or inductive
grounding are different according to levels and areas
Grid level 5 is the MV grid. Theoretical nominal voltages range from 1 kV to 36 kV; in
practice, the MV distribution grid mainly covers the standard range from 10 kV to 20 kV,
with a few exceptions included in the larger range from 5 kV to 24 kV. The rated capacity of
HV/MV transformers, in primary substations, is usually between 16 MVA and 63 MVA. The
MV level is realized as a three-phase delta system consisting mostly of overhead lines in
rural areas and underground cables in suburban areas. Meshed system topologies can be
found in urban areas. In suburban and rural areas often a mixture of open- and closed-loop
structures exist. In rural areas pure branch feeder configurations can be found. Typical
interconnection customers are large industrial facilities as well as utility-scale generation
units with a rated capacity of up to some MW.
The local distribution grid (level 7) supplies small industrial areas and households with LV
(<1 k). Many of the renewable energy power plants such as PV plants are connected to this
grid level. The three-phase LV distribution level has nominal voltage 400 V/230 V. The
rated capacity of Delta-Star MV LV transformers in secondary substations for energy
supply to households, is typically from 250 kVA to 1,000 kVA while higher rates (typically
1600 kVA, 2,500 kVA) are possible for MV industrial users or energy producers. The
nominal line-to-line voltage at LV levels is 400 V 10% (steady-state voltage ranges defined
by EN50160). The system is realized as a three-phase star system with a neutral conductor
comprising mostly of underground cables. In rural areas overhead lines can still be found.
Most of the LV grids consist of branch feeders only. For earthing TN-C systems, as defined
by IEC 60364-1, are most common. Typical interconnection customers are private
households and small commercial buildings.
Most of the Swiss PV capacity can be found at LV levels.
148
Figure 63: Structure of the documents for the liberalized market rules (Source: VSE/AES, Swiss
distribution assessment)
While the laws and ordinances are directly made by federal council, the most important
technical-regulatory documents giving technical details about various mechanisms and
performances required are summarized below:
MMEE CH: Market Model for electrical energy (Source and author: AES/VSE Current
version: 2013). This document mainly gives the description of the market model, with
details about all contractual relations required between each various electricity market
actors, roles and responsibilities, etc.
MUR T/ NNMUe: Transmission grid operation model (Source and main author: Swiss
grid with AES/VSE collaboration). This document mainly describes legal relations with
transmission grid, physical description of transmission grid, description of the
transmission grid role and responsibilities, fares and tariffs related to electricity flux
with transmission grid, etc.
TC: Transmission Code (Source and author: Swiss Grid).This document has the most
detailed requirements in terms of allowed static voltage fluctuations, PQ requirements,
U/f chart, P/f chart, frequency control from Germany, FRT (Sources: chapter 6.5.2;
6.5.5) among other legal wordings concerning grid operation ancillary services and grid
connection to transmission grid. Requirements for power plants' grid connection are
applicable to 50 kV and above. This document will be referred below as TC 2010
(current version).
149
Distribution Code: Distribution Code (Source and author: AES/VSE). This document
defines technical principles for grid connection and operation for distribution grids in
Switzerland (it has to be mentioned, however, that much technical information is fully
delegated to local bilateral connection contracts) (e.g., position of the PCC, minimum
and maximum permanent voltages, etc .). This document contains very little technical
information relevant to PV (e.g., no UfPt, PQ chart, FRT, etc.). Requirements for power
plants' grid connection are applicable to 50 kV and above. This document will be
referred below as DC 2011 (current version).
There is no specific code with power system technical requirements and legal binding
force that was made so far made specifically for PV.
The following documents must, however, be mentioned:
ESTI Nr 233.0710 Solar PV: This Swiss Norm is mainly containing reference to
applicable other norms (e.g., EN) and existing legislative documents applicable. It
focuses on safety, EMC, and power quality (e.g., earthing and neutral concepts and
related safety measures, reference to EN norms for harmonics, etc.).
Important remark about ENTSOE draft grid code current work and its relation with
Switzerland:
Some of the most impactful and most detailed technical grid code ever made to be
applicable for PV and other renewable sources is currently ongoing in Brussels for
application throughout Europe, via the recent current network codes developments (e.g.,
Requirement for generators from the connection codes package, 24 January 2012, which
will be named ENTSOE
in this document).
Even though the Swiss grid is a member of ENTSO-E, gathering 41 TSOs from 34 countries
in Europe, the Switzerland federal state has special status in relation with this project, and
all technical requirements will have to be subject to Swiss laws to become legally binding
and applicable in Switzerland (Source: Network code for requirement for grid connection
applicable to all generators, FAQ, 24 January 2012, chapter 2, ENTSOE internet site). The
application of ENTSOE 2012 technical requirements to PV and other electrical power
plants is therefore uncertain at the time of this report.
ENTSOE will become legally binding to all countries in Europe and it will have direct
impact to all TSOs, DSOs, PV manufacturers, and the whole electrical industry in general.
150
Review Answer 7 of the FAQ to learn more about the European Union position concerning
the relation between domestic PV and the transmission grid, and Answer 3 of the FAQ to
learn about the Switzerland case (Source: ENTSOE internet site, Network code for
requirements for grid connection applicable to all generators Frequently Asked Questions,
24 January 2012).
151
1550
111
28
1,300 kW
5,200 MWh
2 x 1,000 kVA
230 V / 400 V
155
Simulation method
As most of the loads in the grid are domestic households, the loads are modeled using highresolution load profiles (Source: Distribution Grid Analysis and Simulation with
Photovoltaics (DiGASP), Christof Bucher et al., PV+Grid, PV ERA NET). The output of the PV
systems is modeled using stochastic irradiance profiles developed in the same project as
the load profiles.
The load flow computation is done using MATLAB and the Matpower toolbox. The temporal
resolution of the simulation was chosen between 1 and 15 minutes, depending on the type
of simulation. The grid is modeled using the NEPLAN model of ewz, which takes into
account both the impedance and the reactance of the cables. For different grid integration
measures, the PV hosting capacity of the grid was computed. The grid voltage rise was
thereby the major restriction for further increase of the PV penetration.
To compute the PV hosting capacity of the grid, all roofs are homogeneously equipped with
PV systems. The size of the systems is chosen proportionally to the roof surfaces and
gradually increased until the voltage at any point in the grid exceeded the permitted
voltage tolerance band. The same procedure is applied to find the thermal current
limitation of the grid.
Results
The results are summarized in Table 47.
Table 47: Results Summary of the Case Study
Simulation
Method
DACHCZ: no
measures
Correlation with
load
RPC: Reactive
power control
Storage
Remark
Result
156
DSM: Demand
side management
Figure 68 shows the voltage histogram for one year, using a temporal resolution of 5
minutes. The first voltage violation occurs between scenario 3 (PV pen. = 13.4%) and
scenario 4 (PV pen. = 20.1%), while the first line overload occurs only with scenario 6 (PV
pen. = 33.5%).
Figure 68: Voltage histogram and maximum current loading for different PV penetration scenarios
157
Even though PV in Switzerland today represents a small part of the energy mix (0.5% to
1% of the produced electricity), it is growing fast. It is therefore strongly suggested in the
present document to analyze grid needs in anticipation, to observe the requirements from
a panel of countries' grid codes, and to investigate potential opportunities for PV
performance.
Section 2.10.3 above gave the following indications about the potential grid needs and/or
potential capabilities that could be investigated:
Frequency range
Frequency control
Islanding
P control
PQ capabilities:
Steady-state voltage control
Taking into account the very fragmented nature of the distribution grid management
structure in Switzerland (approx. 800 DSOs and one TSO), Switzerland's big challenges will
be, among others, to develop robust, simple, and generic requirements to be relevant both
for the whole grid and for the PV industry.
158
The Tk8 committee from electrosuisse is likely to work on potential evolution of the
DACHZ (Source: internet site of electrosuisse) or potential other norms or standards,
and ESTI is also working on its norm applicable for PV. It can only be assumed that a
working group will be raised in the near future to work on the requirements mentioned in
the present document.
The following two projects are to be mentioned:
Cooperation D-A-CH Smart Grids: The cooperation D-A-CH Smart Grids is based on a
Memorandum of Understanding between the three neighboring countries in
Germany, Austria, and Switzerland. The main goal of this effort is to closely
collaborate in the field of R&D and dissemination in the emerging smart grid area.
To coordinate the work of specific cross-cutting topics, four task forces have been
established.
159
United States of
America
Author:
Barry Mather1
1National
160
Figure 70: Typical simplified electrical power system structure in the United States
Typical distribution circuit nominal operating voltages are 34.5 kV, 33 kV, 24 kV, 13.8 kV,
13.2 kV, 12.47 kV, 12 kV, and 4.16 kV. Many more nominal operating voltages are used by
individual utilities but are typically within the range of voltages shown above. For instance,
a utility may use equipment designed for 12.47 kV nominal operating but will operate with
161
a nominal distribution circuit voltage of 12 kV resulting in a 3.7% lower per unit voltage.
There are approximately 3,000 distribution utilities in the United States consisting of
about 200 investor-owned utilities (IOUs), 840 electrical cooperatives, and the remainder
being municipal utilities [138] and each system is planned and operated in a different
way. This fact, added to the cultural and climatic diversity present across the 50 states that
comprise the United States, makes it is easy to see that there is a lot of variation in the
topology of the distribution circuits. Taking this into account, the following discussion
attempts to present a description of typical distribution system topology.
Typical Distribution Circuit Length and Voltage Level:
Most distribution circuits in the United States are MV circuits with nominal voltage ranges
between 14 kV and 12 kV. Lower voltage circuits (i.e., 4.16 kV) are generally being phased
out and upgraded when possible. Higher voltage circuits (i.e., 34.5 kV) are not common but
are representative of circuits that may have been built relatively recently and likely serve
relatively rural areas (long circuit length) with significant load.
The length of a typical 15 kV Class (i.e., 12.47 kV) distribution circuit depends heavily on
the density of the load it serves as well as the types of loads it serves, such as industrial,
commercial, and residential. In areas with higher load density a circuit may be as short as 3
miles long. In very rural areas and for distribution circuits that serve mainly rural
residential loads, a distribution circuit may be 30 or more miles long. The vast majority of
distribution circuits are bracketed by the two examples given above.
Typical Distribution System Equipment and Controls:
Distribution system equipment often used in the United States includes: load tap changing
transformers, voltage regulators, fixed and controlled capacitor banks, and manual and
remotely controllable pole switches. Load tap changing transformers allow the voltage of
the distribution circuit or circuits supplied by a single substation transformer to be
regulated within a desired tolerance. Line drop compensation (LDC) is often used which
effectively regulates the voltage at a predetermined point on the distribution circuit down
the line from the substation. Voltage regulators operate in much the same was as load tap
changing transformers except that they can be located anywhere on the distribution circuit.
Voltage regulators have controllers very similar to load tap changing transformers and
often implement LDC as well. Capacitor banks are used by utilities to supply some or all of
the reactive power requirements of the loads operating on the distribution circuit. Fixed
capacitor banks are sometimes used if reactive power support and/or voltage support is
needed on the circuit at that specific location at all times of the day. Otherwise, controlled
capacitor banks are employed to connect and disconnect the capacitor from the
distribution circuit when appropriate. Pole switches are used to reconfigure the
distribution circuit and/or are used to isolate sections of the distribution circuit for
maintenance. Some utilities use remotely controllable pole switches to enable automatic or
162
163
Recently, an amendment to IEEE 1547, IEEE P1547a [142], has been under development.
The amendment is considering technical requirement changes for the active regulation of
point of common-coupling voltage, over- and under-voltage trip levels and times, and overand under-frequency trip levels and times.
Regulatory Framework:
Distribution utilities are regulated in a variety of ways in the United States. Investor-owned
utilities (IOUs) are typically regulated by state-level public utility commissions (PUCs). All
IOUs within a state are all regulated by the singular state PUC. Electric cooperatives and
municipal utilities are usually regulated or advised by representatives from their
constituents. While many states have aggressive renewable portfolio standards (RPSs) that
encourage the development of renewable energy resources such as PV energy, it is not a
general requirement that all PV systems must be interconnected. Generally, PV system
installers or owners file a formal request for interconnection, and this request is processed
in a prescribed and regulated manner. Figure 71 shows a simplified flow diagram of the
typical steps in a formal request of interconnection.
Figure 71: Typical simplified flow diagram of the interconnection process of interconnecting DG
164
In the typical interconnection process, as mandated by the utility regulator, there are
usually a number of screens that are applied to the interconnection request to filter out
requests that will have little to no impact on the distribution system to which they
interconnect. These are often based on the total nameplate rating of the generating facility
and the voltage which it interconnects These are often referred to as the fast track
screens )f a project fails one of the fast track screens, a study is completed (usually at the
cost of the PV developer). This study is meant to determine the impact the interconnected
system will have and also to determine any upgrades required to accommodate the
requested interconnection. If upgrades are required, they are to be paid by the PV
developer.
which is also known at the NREL SCE (igh-Penetration PV )ntegration Project [145].
Three other high-penetration PV case studies in the United States are provided in a
condensed format in [146].
Fontana, CA High-Penetration Case Study:
In 2009 SCE began the Solar Photovoltaic Program (SPVP), with the goal of installing 500
MW of primarily rooftop PV in their service territory by 2015. This program focused on
distribution-connected utility-scale (approximately 1 5 MW) PV systems that would be
either utility-owned generation or built and operated by independent power producers.
The first PV system installed as part of the program was a 2 MW PV system located on a
warehouse rooftop near the city of Fontana, California. The distribution circuit with which
the 2-MW PV system is interconnected is shown in Figure 72. This circuit was chosen as
one of three study circuits to be studied under the NREL/SCE High-Penetration PV
Integration Project.
GridSense LT40
on main line after
underground cable
section
GridSense LT40s
on main branch
headed south
GridSense LT40
on main line
after PV plant
4.5 MW PV
system
Substation
GridSense LT40
on main line
outside of
substation
approx. 1 Mile
3-phase NREL DMU
in pad-mount
transformer near end
of line
Figure 72: Circuit diagram of the NREL/SCE High-Penetration PV Integration Project s Fontana,
California, study circuit
As SCE s SPVP project progressed, an additional 2.5 MW (for a total of 4.5 MW) of PV were
installed on the Fontana, CA study circuit in two stages. First, the original 2-MW system
was expanded to 3 MW and later an additional 1.5-MW system was added with a POI very
near the original PV installation. The circuit was instrumented using both commercially
available GridSense LT40 MV line current monitors and an NREL developed distributionlevel voltage phasor measurement unit (PMU) called a distribution monitoring unit (DMU)
[147]. Instruments were placed on the circuit to measure the impacts of high-penetration
PV integration developed by the eventual 4.5 MW of installed PV.
The 4.5 MW PV plant is connected to the local 12 kV distribution system about two miles
from the local substation. The entire length of the interconnected distribution circuit is 7.8
miles including all mainline and branch circuits. The distribution circuit serves primarily
commercial customers, as most of the area served by the circuit is a warehouse district.
Near the end of the distribution circuit other types of commercial loads, such as retail
shops in a mall, are served.
166
Voltage regulation of the circuit is accomplished by switched capacitor banks placed along
the length of the mainline of the distribution circuit. The capacitors are controlled using a
time schedule with a voltage override. The voltage override operating set points are
adjusted automatically if the ambient temperature, measured at the capacitor bank
controller, is above 90 F. Voltage regulation at the substation is also accomplished using
switched capacitors located at the substation. The capacitors at the substation are operated
to both regulate the voltage at the substation bus bars and compensate Var flows in the
subtransmission system.
The circuit consists of an express overhead run of 653 ACSR (336 ACSR on the neutral)
conductors running just past the PV system installation. Then the circuit continues to
additional loads via primarily underground cabling. The mainline cabling is mostly 1000
mil JCN but tapers to 750 CLP near the end of the mainline circuit. Overall, the PV system
being nearly the first interconnection to the distribution circuit after it leaves the
substation, along with the large conductor size, means that the PV system is connected to a
relatively stiff (from a voltage standpoint) interconnection.
The peak load on the interconnected distribution circuit is approximately 4.2 MW, resulting
in a PV penetration of potentially 107%. As the load profile is dominated by commercial
loads, specifically warehouses, the circuit loading on days when the warehouses are not
operating, such as holidays and weekends, the circuit peak load for the day can be much
lower. The peak loading for an off-peak day is estimated to be 2.6 MW. Reverse power flow
along the distribution line to the substation is common during clear days particularly since
the PV system was increased by 2.5 MW.
As part of the analysis completed on the Fontana, California, study circuit an investigation
into the possible mitigation of the voltage related impacts of the existing high-penetration
PV integration was undertaken. Specifically, the use of advanced inverter capabilities such
as off-unity power factor operation, constant Var operation and other POI voltageresponsive Var output capabilities were modeled and analyzed. Figure 73 shows the result
of one such investigation looking at the relationship between potential off-unity power
factor operation, the effect of partial generation variation on a minute-to-minute timescale
due to weather conditions, and the resulting voltage difference seen at the POI of the
operating PV system (in this case the larger 3 MW PV system). As shown in the surface plot
the voltage variation seen at the PV system POI due to the variations in PV system power
output over short time periods can be partially mitigated by the implementation of a
constant off-unity power factor operating set point for the operating PV system. Limiting
these circuit-level voltage variations is important for two reasons. The first is that voltage
variations at the PV system POI will be seen at other customer service POIs along the entire
circuit, particularly customers located on the distribution circuit after the PV system POI.
Secondly, the voltage along the length of the Fontana, California, study circuit is regulated
solely by switched capacitors attached to the MV distribution circuit, and the voltage
variations caused by the interconnected PV system could potentially disrupt the expected
operation of these devices whose control depends partially on the circuit voltage seen at
their locations on the distribution circuit. Further, both of the reasons listed above are
commonly the limiting factor regarding the amount of variable generation, such as PV, that
167
can be accommodated on a distribution circuit without the need for potentially expensive
distribution circuit upgrades. By limiting the amount of voltage variation caused by the
interconnected PV systems, it is very likely that more PV systems could be accommodated
before another PV impact limit is reached.
Figure 73: Surface graph of POI voltage change for a range of operating power factor set points and for
a range of power production loss metrics
As Figure 73 shows, the optimal mitigation strategy (based on a constant power factor set
point in this case) requires the determination of the likely maximum minute-over-minute
power production loss or gain that is expected from the given operating PV system
characteristics. Table 48 shows the results of analysis completed to determine the number
of instances during the year 2011 that the operating PV systems would experience at
minute-over-minute loss or gain of power production of 33%, 50%, or 75%. This analysis
was completed using data derived from weather satellite data and was processed using
cloud motion vector analysis (CPR). The result is that for the aggregated systems on the
Fontana, CA study feeder, that are located in close proximity, there are no instances when
the short-term variation is greater than 75%. At the variation level of 50% the number of
occurrences is on the order of two or three dozen. Instances when the PV plant output
changes by 33% are quite common. Using this data it was then possible to plan for the
relatively rare occurrence of a 50% minute-over-minute power production loss or gain and
use the resulting modeled POI voltage change to select an appropriate PV system power
factor operating set point.
168
Table 48: PV Power Output Variation Expected for Modeled Year 2011
Table 49: Magnitude of PV Systems Point of Interconnection Voltage Change Related to Operating
Power Factor Set Point and Expected Production Power Loss (Red Numbers Indicate Voltage Variation
Higher than 0.7 volts on a 120-V base.)
Power Factor
-0.90 PF
-0.95 PF
1.00 PF
0.95 PF
0.90 PF
50%
-0.18
0.05
0.59
1.06
1.23
60%
-0.22
0.07
0.72
1.28
1.48
70%
-0.25
0.08
0.84
1.50
1.73
80%
-0.28
0.10
0.96
1.72
1.99
90%
-0.31
0.12
1.09
1.94
2.24
100%
-0.33
0.14
1.22
2.16
2.50
169
170
can be seen. At 2 MW, the voltages present at any given time do not necessarily exceed the
circuit s operational limits, but the change in the circuit s operational characteristics i e ,
voltage level) is apparent.
822
820
818
800
802
806
808
812
Reg #1
814
850
816
864
836
860
824
828
810
840
826
830
Reg #2
832
852
834
858
862
842
888
838
844
890
854
856
846
848
Figure 74: Single-line diagram of the IEEE 34 node test feeder showing the location of the two voltage
regulators and two fixed-shunt capacitors
Figure 75: Color graphs of the POI voltage (node 840) for every minute of the year in 2010 with a) no
PV interconnected and b) a 2-MW PV system connected at node 840
Figure 76 shows the results of a further investigation into the modeled operation of the two
voltage regulators for the case of a 1 MW system interconnected at node 840 (end of the
mainline). The modeled power output of the PV plant is shown time-aligned with the tap
positions of the two voltage regulators. As shown, there is a direct correlation to voltage
regulator operation (tapping up or down) and variability in the PV systems power output.
This is clearly shown in the graphs that show the same data for a time period between 10
a.m. and 1 p.m. Additional quasi-static time-series simulations were completed to
determine the locational sensitivity of the PV interconnection location on the impact of the
two voltage regulators. The results of that analysis are presented in Figure 77. This figure
171
shows the percentage increase modeled for both voltage regulators when a 1 MW PV
system is interconnected at any node along the length of the mainline distribution circuit.
This type of analysis provides considerable insight to distribution engineers tasked with
evaluating the impact of requested PV interconnections on their distribution system. As
shown, the impact on voltage regulator operation is minimal if the PV system is
interconnected within about 7 miles of the substation. Voltage regulator #2 is not heavily
impacted unless the PV is interconnected past voltage regulator #2 (about 20 miles from
the substation in this case).
Figure 76: Model results of voltage regulator operation during periods of highly variable generation on
June 16, 2010. The graphs on the right show regulator operation during the entire day. The graphs on
the left show operation from 10 a.m. to 1 p.m.
172
Figure 77: The modeled voltage regulator operation impact of interconnecting 1 MW of PV at any
point along the distribution circuit
173
174
R
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
175
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
Alice Springs News Online. Alice Solar City - Over to you. 2013 13
November 2013]; Available from:
http://www.alicespringsnews.com.au/2013/06/18/solar-cityover-to-you/.
176
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
177
[41]
[42]
[43]
[44]
[45]
[46]
[47]
[48]
[49]
178
[50]
MetaPV Project-Website
http://www.metapv.eu/
[51]
[52]
[53]
[54]
[55]
[56]
[57]
People's Republic of China standard GB / T 293192012: Technical requirements for connecting photovoltaic power
system to distribution network , June 1, 2013 implementation.
[58]
[59]
[60]
[61]
[62]
People's Republic of China Energy Industry Standard NB / T 320042013: Technical specification of grid-connected PV inverter, August
1, 2013 implementation.
179
[63]
Sun Honghang, Zhi Qiang, Wang Yibo, and et al. Chinas solar
photovoltaic industry development: The status quo, problems and
approaches. Applied Energy 118 (2014) 221230.
[64]
[65]
[66]
[67]
[68]
[69]
Association of German Network Operators VDN (2004); EEGErzeugungsanlagen am Hoch- und Hchstspannungsnetz Berlin
http://www.vde.com/de/fnn/dokumente/documents/rl_eeg_hh_
vdn2004-08.pdf
[70]
[71]
[72]
bertragungsnetzbetreiber Berlin
http://www.netzentwicklungsplan.de/system/files/documents/T
ransmissioncode%20_2007_UeNB.pdf
[73]
[74]
[75]
[76]
[77]
[78]
[79]
[80]
[81]
181
[83]
[84]
[85]
[86]
[87]
[88]
Idlbi B., Diwold K., Stetz T., Wang H., Braun M.: Cost-Benefit
Analysis of Central and Local Voltage Control Provided by
Distributed Generators in MV Networks, Powertech Conference,
Grenoble, 2013
[89]
[90]
[91]
182
[92]
PV Symphonie Zwischenbericht
[93]
[94]
[95]
[96]
[97]
[98]
[99]
A. Gigantidou (2012); RES at Crete, Anemologia, Issue SeptemberDecember 2012, Vol. 75, page 30-33, (in Greek) and private
communication.
[100]
[101]
[102]
183
[103]
[104]
[105]
[106]
[107]
Annual Report of the Italian Autority for Electric Energy and Gas
AEEG (2013 March 31st) RELAZIONE ANNUALE SULLO STATO
DE) SERV)Z) E SULL ATT)V)T SVOLTA Available at web address:
http://www.autorita.energia.it/allegati/relaz_ann/13/RAVolumeI
_2013.pdf.
[108]
CEI standard number 0-16 (2013) Reference technical rules for the
connection of active and passive consumers to the HV and MV
electrical networks of distribution Company Available ate the web
address: http://www.ceiweb.it/it/comunicati/news/331-normacei-0-16-edizione-dicembre-2012-con-aggiornamento-maggio2013.html.
[109]
CEI standard number 0-21 (2012) Reference technical rules for the
connection of active and passive users to the LV electrical Utilities
Available at the web address
http://www.ceiweb.it/it/comunicati/news/234-norma-cei-0-21aggiornamento-dicembre-2012.html.
[110]
[111]
The photovoltaic
http://www.autorita.energia.it/it/docs/08/099-08arg.htm.
[112]
[113]
[114]
[115]
[116]
[117]
[118]
Politecnico di Milano
Impatto della generazione diffusa
sulle reti di distribuzione BT esiti dell analisi su un campione
ridotto LV study available at web address
185
http://www.autorita.energia.it/allegati/docs/10/22310argall.pdf (allegato1.pdf).
[119]
[120]
[121]
[122]
[123]
[124]
[125]
[126]
[127]
[128]
186
[130]
[131]
[132]
[133]
[134]
[135]
[136]
[137]
[138]
[139]
187
[140]
[141]
[142]
[143]
[144]
[145]
[146]
[147]
[148]
[149]
188
[150]
[151]
[152]
[153]
189
A
Table 50: Applied Local Control Strategies in German Case Study [84]
Category
Scenario
Applied Parameterization
Grid
Reinforcement
On-load tap
changer of
MV/LV
transformer
Autonomous
Inverter Control
Strategies
No Voltage Control
CosPhi(P)
Q(V) with PF limitation
Q(V) without PF limitation
Figure 78: Applied Q(V)/P(V) characteristic in the German case study [84]
190